DESC:CROSS REFERENCE TO RELATED APPLICATION
This Application is based upon and derives the benefit of Indian Provisional Application Number 202441009447 filed on 12th February 2024, the contents of which are incorporated herein by reference.

TECHNICAL FIELD
[0001] Embodiments of the present disclosure generally relate to artificial intelligence systems for offering hyper-personalized services. More particularly, the present disclosure relates to systems and methods for managing behaviour-based personalized services for end users in a computing environment.
BACKGROUND
[0002] Generally, increasing complexity of user behaviour and the rising demand for personalized services have posed significant challenges for service providers. As users engage with digital platforms across diverse domains, their expectations for highly customized experiences have intensified. However, conventional methods of user categorization and recommendation often fail to capture nuances of individual behaviour, leading to generic services that does not align with user interests, preferences, or expectations. This misalignment results in poor user satisfaction, reduced engagement, and diminished effectiveness of service offerings.
[0003] Behaviour-driven personalization has emerged as a key approach for enhancing user experiences and optimizing service delivery. By leveraging behavioural insights, the service providers may create highly engaging and tailored experiences that drive active user participation and long-term retention. Furthermore, behaviour-based personalization enables targeted service allocation, ensuring that services reach the most relevant users, thereby improving efficiency and increasing return on investment (ROI). However, despite these advantages, existing behaviour-based personalization systems face several technical challenges and limitations.
[0004] One of the primary challenges is the reliance on simplistic user models and limited data sources. Traditional recommendation systems often fail to construct accurate and dynamic user profiles due to their dependence on static, predefined parameters. Such models struggle to adapt to evolving user behaviour, leading to outdated or irrelevant recommendations. Additionally, many existing solutions lack mechanisms for real-time data integration, preventing them from delivering immediate and contextually relevant recommendations based on the latest user interactions.
[0005] Data privacy and security concerns represent another critical challenge. Behaviour-based personalization requires the collection, storage, and processing of vast amounts of user data, including browsing history, interaction patterns, and contextual preferences. However, users are increasingly apprehensive about sharing sensitive information due to concerns over data misuse, unauthorized access, and lack of transparency in data handling practices. Many conventional systems do not incorporate robust privacy-preserving mechanisms, making them susceptible to data breaches, unauthorized profiling, and non-compliance with data protection regulations such as General Data Protection Regulation (GDPR) and the like.
[0006] Another significant limitation is the inability to handle real-time behavioural changes. User preferences and behaviours are inherently dynamic, influenced by factors such as time, context, and external trends. However, many existing personalization models rely on historical data without effectively integrating recent behavioural shifts. This leads to outdated recommendations that fail to capture emerging user interests, reducing the relevance and effectiveness of personalized services.
[0007] Algorithmic bias and fairness concerns also pose a major obstacle in behaviour-based personalization. Many recommendation algorithms are trained on biased datasets, leading to skewed or discriminatory outcomes. Bias in training data may result in unfair treatment of specific user groups, limiting diversity in recommendations and reinforcing existing stereotypes. Furthermore, the lack of explainability in many AI-driven recommendation models makes it difficult to detect and correct biases, reducing user trust and regulatory compliance.
[0008] Scalability and computational efficiency present additional technical hurdles. As service providers scale to accommodate larger user bases, the computational demands of behaviour-based personalization systems increase exponentially. Many existing solutions struggle with high processing overheads, leading to slower response times and increased infrastructure costs. Efficient data processing techniques, optimized machine learning models, and scalable architectures are necessary to address these performance bottlenecks.
[0009] Furthermore, cross-platform and multi-device synchronization challenges hinder seamless personalization. Users interact with digital services across multiple devices and platforms, yet many existing personalization systems fail to provide a unified experience across different touchpoints. The lack of a standardized framework for data interoperability results in fragmented user experiences and inconsistent recommendations across devices.
[0010] Given these challenges, there is a critical need for improved systems and methods for behaviour-based personalized services that enhance accuracy, adaptability, and privacy while overcoming the limitations of existing approaches.
[0011] Therefore, there is a need in the art to provide improved systems and methods for managing behaviour-based personalized services for an end user based on user behaviour, to address at least the aforementioned issues in the prior arts.
SUMMARY
[0012] This summary is provided to introduce a selection of concepts, in a simple manner, which is further described in the detailed description of the disclosure. This summary is neither intended to identify key or essential inventive concepts of the subject matter nor to determine the scope of the disclosure.
[0013] In accordance with an embodiment of the present disclosure, systems for managing behaviour-based personalized services for end users in a computing environment is disclosed. The system periodically obtains a multi-modal data associated with a user from a plurality of data sources. The multi-modal data includes a behavioural data , an emotional data, a demographic data, and an ethnographic data. The system further determines user emotions and behavioural attributes associated with the user based on the obtained multi-modal data using artificial-intelligence model. The behavioural attributes includes a user interaction frequency, time spent on features, and content preferences. The system further generates a user persona model based on the determined user emotions and the behavioural attributes associated with the user. The user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user. Further, the system computes a behavioural score for the user based on the generated user persona model. The behavioural score corresponds to a unique value assigned to the user. The system furthermore predicts one or more probable interactions of the user with one or more applications based on computed behavioural score using generative artificial intelligence (Gen-AI) models. The system further determines one or more modifications to be made to service configurations of the one or more applications by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications. The one or more virtual applications correspond to computer simulated version of the one or more applications. The system further modifies the service configurations of the one or more applications based on the determined one or more modifications. The service configurations comprises at least one of a network throughput, a response time, and inter-service links. Furthermore, the system outputs the computed behavioural score as a token to the one or more applications upon modifying the service configurations of the one or more applications.
[0014] Further, in accordance with an embodiment of the present disclosure, a method for managing behaviour-based personalized services for end users in a computing environment is disclosed. The method includes periodically obtaining a multi-modal data associated with a user from a plurality of data sources. The multi-modal data comprises a behavioural data , an emotional data, a demographic data, and an ethnographic data. The method includes determining user emotions and behavioural attributes associated with the user based on the obtained multi-modal data using artificial-intelligence model. The behavioural attributes includes a user interaction frequency, time spent on features, and content preferences. The method further includes generating a user persona model based on the determined user emotions and the behavioural attributes associated with the user. The user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user. Furthermore, the method includes computing a behavioural score for the user based on the generated user persona model. The behavioural score corresponds to a unique value assigned to the user. The method further includes predicting one or more probable interactions of the user with one or more applications based on computed behavioural score using generative artificial intelligence (Gen-AI) models.
[0015] Also, the method includes determining, one or more modifications to be made to service configurations of the one or more applications by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications. The one or more virtual applications correspond to computer simulated version of the one or more applications. Additionally, the method includes modifying the service configurations of the one or more applications based on the determined one or more modifications. The service configurations includes at least one of a network throughput, a response time, and inter-service links. Moreover, the method includes outputting the computed behavioural score as a token to the one or more applications upon modifying the service configurations of the one or more applications.
[0016] To further clarify the advantages and features of the present disclosure, a more particular description of the disclosure will follow by reference to specific embodiments thereof, which are illustrated in the appended figures. It is to be appreciated that these figures depict only typical embodiments of the disclosure and are therefore not to be considered limiting in scope. The disclosure will be described and explained with additional specificity and detail with the appended figures.
BRIEF DESCRIPTION OF DRAWINGS
[0017] The disclosure will be described and explained with additional specificity and detail with the accompanying figures in which:
[0018] FIG. 1A-B illustrates a block diagram representation of an exemplary computing environment for managing behaviour-based personalized services for end users, in accordance with an embodiment of the present disclosure;
[0019] FIG. 2 illustrates a block diagram representation of an exemplary computing system, such as those shown in FIG. 1, for managing behaviour-based personalized services for end users, in accordance with an embodiment of the present disclosure.
[0020] FIG. 3 illustrates an example block diagram of a computing system such as those shown in FIG. 1, capable of managing behaviour-based personalized services for end users, in accordance with an embodiment of the present disclosure;
[0021] FIG. 4 illustrates an example graphical user interface depicting a personality assessment or decision-making tool using different types of hats for personality assessment, in accordance with an embodiment of the present disclosure;
[0022] FIG. 5 illustrates an example graphical user interface depicting chatbot conversations for managing behaviour-based personalized, in accordance with an embodiment of the present disclosure;
[0023] FIG. 6 illustrates an example blockchain network depicting a process of transmission and validation of a user token for providing personalized services for end users, in accordance with an embodiment of the present disclosure; and
[0024] FIG. 7 illustrates a process flow diagram representation of an exemplary method for managing behaviour-based personalized services for end users, in accordance with an embodiment of the present disclosure.
[0025] Further, those skilled in the art will appreciate that elements in the figures are illustrated for simplicity and may not have necessarily been drawn to scale. Furthermore, in terms of the construction of the device, one or more components of the device may have been represented in the figures by conventional symbols, and the figures may show only those specific details that are pertinent to understanding the embodiments of the present disclosure so as not to obscure the figures with details that will be readily apparent to those skilled in the art having the benefit of the description herein.
DETAILED DESCRIPTION OF THE DISCLOSURE
[0026] For the purpose of promoting an understanding of the principles of the disclosure, reference will now be made to the embodiment illustrated in the figures and specific language will be used to describe them. It will nevertheless be understood that no limitation of the scope of the disclosure is thereby intended. Such alterations and further modifications in the illustrated system, and such further applications of the principles of the disclosure as would normally occur to those skilled in the art are to be construed as being within the scope of the present disclosure. It will be understood by those skilled in the art that the foregoing general description and the following detailed description are exemplary and explanatory of the disclosure and are not intended to be restrictive thereof.
[0027] In the present document, the word “exemplary” is used herein to mean “serving as an example, instance, or illustration.” Any embodiment or implementation of the present subject matter described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.
[0028] The terms “comprise,” “comprising,” or any other variations thereof, are intended to cover a non-exclusive inclusion, such that one or more devices or sub-systems or elements or structures or components preceded by “comprises…. a” does not, without more constraints, preclude the existence of other devices, sub-systems, additional sub-modules. Appearances of the phrase “in an embodiment,” “in another embodiment” and similar language throughout this specification may, but not necessarily do, all refer to the same embodiment.
[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this disclosure belongs. The system, methods, and examples provided herein are only illustrative and not intended to be limiting.
[0030] A computer system (standalone, client or server computer system) is configured by an application may constitute a “module” (or “subsystem”) that is configured and operated to perform certain operations. In one embodiment, the “module” or “subsystem” may be implemented mechanically or electronically, so a module includes dedicated circuitry or logic that is permanently configured (within a special-purpose processor) to perform certain operations. In another embodiment, a “module” or “subsystem” may also comprise programmable logic or circuitry (as encompassed within a general-purpose processor or other programmable processor) that is temporarily configured by software to perform certain operations.
[0031] Accordingly, the term “module” or “subsystem” should be understood to encompass a tangible entity, be that an entity that is physically constructed permanently configured (hardwired) or temporarily configured (programmed) to operate in a certain manner and/or to perform certain operations described herein.
[0032] Embodiments of the present disclosure may include systems and methods for managing behaviour-based personalized services for end users in a computing environment is disclosed. The system periodically obtains a multi-modal data associated with a user from a plurality of data sources. The multi-modal data includes a behavioural data, an emotional data, a demographic data, and an ethnographic data. The system further determines user emotions and behavioural attributes associated with the user based on the obtained multi-modal data using artificial-intelligence model. The behavioural attributes includes a user interaction frequency, time spent on features, and content preferences. The system further generates a user persona model based on the determined user emotions and the behavioural attributes associated with the user. The user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user. Further, the system computes a behavioural score for the user based on the generated user persona model. The behavioural score corresponds to a unique value assigned to the user. The system furthermore predicts one or more probable interactions of the user with one or more applications based on computed behavioural score using generative artificial intelligence (Gen-AI) models. The system further determines one or more modifications to be made to service configurations of the one or more applications by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications. The one or more virtual applications correspond to computer simulated version of the one or more applications. The system further modifies the service configurations of the one or more applications based on the determined one or more modifications. The service configurations comprises at least one of a network throughput, a response time, and inter-service links. Furthermore, the system outputs the computed behavioural score as a token to the one or more applications upon modifying the service configurations of the one or more applications.
[0033] In an aspect, the system may be configured to compute a micro-segmentation of the end user through one or more behavioural identifiers for providing targeted personalized services. The system may include a user profiling model which may be a nonlinear pre profiling model representing diverse human emotions and behavioural attributes associated with the end user. Human emotions and behavioural attributes may be determined based on attitudinal interactions, tactile information such as for example, but not limited to, demographic and ethnographic and persona exhibiting different behaviours in different contexts of the end user. Further, the system may be configured to generate a unique behavioural ID by computing a unique value and attaching the unique value with the unique behavioural ID. In one example embodiment, the behavioural ID may be a chromatic Identity (ID) or any other form of representation of behavioural information. The unique behavioural ID may be for example, a Non-Fungible Token (NFT) that may be for ensuring transmissibility across various networks, and multi-variant systems. The unique behavioural ID may be able to preserve details, preferences, and behavioural patterns of the end user. In an aspect of the present disclosure, the system may be configured to apply data encryption techniques to maintain security and confidentiality of the details, preferences, and behavioural patterns of the end user. Further, the system may be configured to provide Quality of Services (QoS) such as hyper personalized/hyper-identified or customized services to the end user based on the unique behavioural ID associated with the end user. Furthermore, the system may manage recommendations of services to the end user based on the unique behavioural ID associated with the end user. The unique behavioural ID may be for example, the Non-Fungible Token (NFT) which maintains complete privacy of behavioural attributes, and therefore other systems and network providing services may not be fully aware of the actual behavioural attributes of the end user.
[0034] Referring now to the drawings, and more particularly to FIGs. 1A-B through FIG. 7, where similar reference characters denote corresponding features consistently throughout the figures, there are shown preferred embodiments, and these embodiments are described in the context of the following exemplary system and/or method.
[0035] FIG. 1A illustrates a block diagram representation of an exemplary computing environment for managing behaviour-based personalized services for end users, in accordance with an embodiment of the present disclosure.
[0036] According to FIG. 1A, the computing environment 100a for managing behaviour-based personalized services for end users may include one or more user devices 106A-N (collectively referred herein as one or more user devices 106) associated with one or more end users. The one or more user devices 106 are communicatively coupled to a computing system 102 via a network 104. Further, the computing environment 100a may include a plurality of applications 108A-N connected to the computing system 102 via the network 104. Further, the computing environment 100a may include a plurality of external databases 114A-N (also referred herein as data sources 114A-N) communicatively coupled to the computing system 102 and the one or more user devices 106 via the network 104.
[0037] In an exemplary embodiment of the present disclosure, the one or more user devices 106 may include, but not limited to, smart phones, laptop computers, desktop computers, tablet computers, wearable devices, smart watches, and the like. Further, the one or more end users may include, but not limited to, any individuals or humans. Further, the computing system 102 may be accessed by the one or more users from the one or more user devices 106 by using for example, but not limited to, application interface 112A-N such as for example, application plug ins or application programming interfaces, or the like.
[0038] Further, the network 104 may include, but not limited to, physical networks such as Local Area Network (LAN), Wide Area Network (WAN), Metropolitan Area Network (MAN), Personal Area Network (PAN), wireless networks such as Cellular Networks, Satellite Networks, Wireless Fidelity (Wi-Fi) Networks, and the like.
[0039] Further, the plurality of external databases 114A-N may be configured to store, manage, organize, distribute, and protect data relevant to the computing system 102. Furthermore, the computing system 102 may include plurality of modules 110 configured to be executed by one or more hardware processors (not shown). Further, the computing environment 100a may include plurality of external databases 114A-N (collectively referred herein as one or more external data sources 114A-N) communicatively coupled to, but not limited to, the one or more user devices 106, via the network 104. Further, the one or more external data sources 114A-N may include, but not limited to, user interaction data sources, which encompass web browsing history, mobile app usage logs, social media activity, chatbot conversations, and online shopping behaviour, transactional data sources, which cover e-commerce purchase records, banking and financial transactions, digital wallet activity, and subscription or membership services and the like. Demographic and ethnographic data sources may be obtained from government census data, publicly available databases, and employment or education records. Environmental and contextual data sources include GPS and location tracking systems, weather APIs, IoT-enabled smart home devices, and traffic monitoring systems. The one or more external data sources 114A-N may include, healthcare and wellness data sources, enterprise and workforce data sources and the like.
[0040] In some embodiments, the computing environment 100a may include a cloud interface, cloud hardware and OS, a cloud computing platform, and a database. The cloud interface enables communication between the cloud computing platform and the user device 106. Also, the cloud interface enables communication between the cloud computing platform and the web application. The cloud hardware and OS may include one or more servers on which an operating system is installed and including one or more processing units, one or more storage devices for storing data, and other peripherals required for providing cloud computing functionality. The cloud computing platform is a platform which implements functionalities such as data storage, data analysis, data processing, data communication on the cloud hardware and OS via APIs and algorithms and delivers the aforementioned cloud services. The cloud computing platform may include a combination of dedicated hardware and software built on top of the cloud hardware and OS. As used herein, “cloud computing environment” refers to a processing environment comprising configurable computing physical and logical assets, for example, networks, servers, storage, applications, services, and the like and data distributed over the cloud platform. The cloud computing environment or the computing environment 100a provides on -demand network access to a shared pool of the configurable computing physical and logical assets. The server may include one or more servers on which the OS is installed. The servers may comprise one or more processors, one or more storage devices, such as, memory units, for storing data and machine -readable instructions for example, applications and application programming interfaces (APIs), and other peripherals required for providing cloud computing functionality.
[0041] In some embodiments, the computing system 102 (also referred herein as “system” 102) may be a remote server, a web server, an edge server, or a blockchain node in a blockchain network.
[0042] The user device 106 may access web applications 108A-B via a an application interface 112A. As an example, a user of a user device 106 may access a particular web application 108 by launching a web browser, such as local web browser, typing into the web browser's address bar a Uniform Resource Locator (URL) address for a web page whose rendering causes execution of the web application 108, and selecting an "enter" key on the user's keyboard. The local web browser may send a Hypertext Transfer Protocol (HTTP) request over the internet to the computing system 102 for resources that correspond to the URL.
[0043] In response to the HTTP request, the local web browser may receive a set of resources that the computing system 102 identified as relevant for the URL (e.g., HTML for a web page, a CSS document, and a JavaScript file). The local web browser may execute the resources, for example, by rendering a parent HTML file and executing other resources referenced therein. The execution of the resources may cause the local web browser to effectively "display" the web application 108 on a display device of the user device 106.
[0044] The web application 108 may be a normal website that includes extra metadata that is installed as part of the browser application. Installable web apps may use standard web technologies for server-side and client-side code. Examples of web applications 108 may include, but not limited to, office applications, games, e-commerce platforms, online shopping platforms, health care platforms, EdTech platforms, social platforms, photo editors, and the like. In an embodiment, the web applications 108 may be deployed on the cloud computing system 102 or on any external enterprise data centre.
[0045] In an example embodiment, the plurality of applications 108A-N may be other types of applications such as local applications, or native applications, cloud applications, and the like.
[0046] In an example embodiment, the plurality of applications 108A-N may be hosted within the user devices 106A-N. In such a case, the plurality of applications 108A-N may be accessed by the user devices 106A-N using the application interface 112A-N.
[0047] In an example embodiment, the computing system 102 is configured to periodically obtain a multi-modal data associated with a user from a plurality of data sources 114A-N. The multi-modal data includes a behavioural data, an emotional data, a demographic data, and an ethnographic data. The computing system 102 is configured to further determine user emotions and behavioural attributes associated with the user based on the obtained multi-modal data using artificial-intelligence model. The behavioural attributes includes a user interaction frequency, time spent on features, and content preferences. The computing system 102 is configured to further generate a user persona model based on the determined user emotions and the behavioural attributes associated with the user. The user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user.
[0048] Further, the computing system 102 is configured to compute a behavioural score for the user based on the generated user persona model. The behavioural score corresponds to a unique value assigned to the user. The computing system 102 is configured to predict one or more probable interactions of the user with one or more applications 108A-N based on computed behavioural score using generative artificial intelligence (Gen-AI) models. The computing system 102 is configured to determine one or more modifications to be made to service configurations of the one or more applications 108A-N by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications. The one or more virtual applications correspond to computer simulated version of the one or more applications 108A-N. The computing system 102 is configured to modify the service configurations of the one or more applications 108A-N based on the determined one or more modifications. The service configurations comprises at least one of a network throughput, a response time, and inter-service links. In an alternate embodiment, the service configurations may also include a higher loyalty points and dedicated customer support, premium access to exclusive features or the like. Furthermore, the computing system 102 is configured to output the computed behavioural score as a token to the one or more applications 108A-N upon modifying the service configurations of the one or more applications 108A-N.
[0049] Those of ordinary skilled in the art will appreciate that the hardware depicted in FIG .1A may vary for particular implementations. For example, other peripheral devices such as an optical disk drive and the like, Local Area Network (LAN), Wide Area Network (WAN), Wireless (for example, Wi-Fi) adapter, graphics adapter, disk controller, input/output (1/0) adapter also may be used in addition or in place of the hardware depicted. The depicted example is provided for the purpose of explanation only and is not meant to imply architectural limitations with respect to the present disclosure.
[0050] Those skilled in the art will recognize that, for simplicity and clarity, the full structure and operation of all data processing systems suitable for use with the present disclosure is not being depicted or described herein. Instead, only so much of a computing system 102 as is unique to the present disclosure or necessary for an understanding of the present disclosure is depicted and described. The remainder of the construction and operation of the computing system 102 may conform to any of the various current implementation and practices known in the art.
[0051] FIG. 1B illustrates a block diagram representation of an exemplary computing environment 100b for managing behaviour-based personalized services for end users 116, in accordance with an embodiment of the present disclosure.
[0052] The computing system 102 for managing behaviour-based personalized services for an end user 116 based on user behaviour may be connected to one or more user devices 106-1, 106-2, 106-3, …, 106-N (individually referred to as the user device 106, and collectively referred to as the user devices 106) via the network 104. In an embodiment, the network 106 may include, by way of example but not limitation, at least a portion of one or more networks having one or more nodes that transmit, receive, forward, generate, buffer, store, route, switch, process, or a combination thereof, one or more messages, packets, signals, waves, voltage or current levels, some combination thereof, or the like. The network 104 may also include, by way of example but not limitation, one or more of a wireless network, a wired network, an internet, an intranet, a public network, a private network, a packet-switched network, a circuit-switched network, an ad hoc network, an infrastructure network, a Public-Switched Telephone Network (PSTN), a cable network, a cellular network, a satellite network, a fiber optic network, and combination thereof.
[0053] The user devices 106A-N may be interchangeably specified as a User Equipment (UE) and may be operated by one or more users 116-1, 116-2, 116-3, …, 116-N (individually referred to as the user 116, and collectively referred to as the users 116). In an embodiment, the user devices 106A-N may include, but not limited to, a mobile, a laptop, and the like. Further, the user devices 106A-N may include, but not limited to, a smartphone, Virtual Reality (VR) devices, Augmented Reality (AR) devices, a general-purpose computer, a desktop, a personal digital assistant, a tablet computer, a wearable device, a metaverse-based device, a mainframe computer, and the like. Additionally, input devices for receiving input from the user device 106 such as for example, a touch pad, touch-enabled screen, electronic pen, and the like may be used. A person of ordinary skill in the art will appreciate that the user devices 106A-N may not be restricted to the mentioned devices and various other devices may be used.
[0054] In an embodiment of the present disclosure, the computing system 102 may include a processor (not shown in FIG. 1B) coupled to a memory (not shown in FIG. 1B) that includes processor-executable instructions, which on execution, causes the processor to execute a sequence of steps described below.
[0055] In an embodiment, the computing system 102 may be configured to provide a framework for learning behavioural identifiers for computing the micro-segment of one or more end users 116. The framework may utilize a nonlinear profiling model representing diverse human emotion and behavioural attributes. An outcome of the framework drives identification and micro-segmentation of the one or more end users 116 through a combination of attitudinal interactions as well as demographic and ethnographic factors.
[0056] The computing system 102 may be configured to generate the unique behavioural ID by identifying persona and understanding, behavioural modalities, and interactions of the one or more end users 116. This may be achieved based on persona exhibiting different behaviours in different contexts of the one or more end users 116. Once the computing system 102 identifies behaviour attributes and emotional attributes, the unique value may be computed and attached to the unique behavioural ID. The value along with behavioural patterns of the end user 116 is a unique way in which the system may provide hyper-personalized/hyper-identified services. Further, the computing system 102 may be configured to provide hyper personalized/hyper-identified services to the one or more end users 116 based on the behavioural ID of the one or more end users 116 including, but not limited to, time spent, user interactions, user emotions, service context, and the like. In an example embodiment, the one or more end users 116 may be exploring a desired property for themselves and ensure to buy the right property based on emotions and behaviours. The framework may create the micro-segmentation of the one or more end users 116 which may be driven by the unique behavioural ID and tactile data including demographic and ethnographic information.
[0057] In an embodiment, the computing system 102 may be configured to implement the unique behavioural ID as a Non-Fungible Token (NFT) to ensure transmissibility across various networks, multi-variant systems and outside the computing system 102. By way of immutability, the unique behavioural ID may be able to preserve details, preferences, and behaviours with complete privacy. This confidentiality may provide the one or more end users 116 with full control over utilization and transmission of the unique behavioural ID based on service context and respective consent the one or more end users 116.
[0058] In another embodiment, the unique behavioural ID may be used to deliver the Quality of Service (QoS) dynamically, by way of understanding needs of the one or more end users 116 through behavioural and emotional attributes. For example, the one or more end users 116 may receive a different level of QoS (such as a high network throughput) to meet the requirements of elevated behavioural impact (impatient or anxiety arising out of high transaction value). The unique behavioural ID may bring enhanced capability to any existing service end points (such as an Application Programming Interface (API)), by injecting non-fungible and fully private behavioural attributes to enable deep hyper-personalization and bespoke humanized experiences.
[0059] In an example embodiment, the service provider in an e-commerce domain may use the unique behavioural ID to provide deeply personalized recommendations for product searches and unique offers/discounts with respect to specific users 116.
[0060] In an example embodiment, the computing system 102 may be configured to constantly learn the user behaviours at various platforms/systems and update the behavioural data accordingly. In one embodiment, the computing system 102 may predict the behavioural ID based on implicit behavioural data including, for example, a historical user data. In an alternate embodiment, the computing system 102 may predict the behavioural ID based on explicit behavioural data which has been recorded at real-time. Using these behavioural data, the computing system 102 may explore new behavioural data applicable to the particular user at a given environment.
[0061] Although FIG. 1B shows exemplary components of the network architecture, in other embodiments, the network architecture may include fewer components, different components, differently arranged components, or additional functional components than depicted in FIG. 1B. Additionally, or alternatively, one or more components of the network architecture may perform functions described as being performed by one or more other components of the network architecture.
[0062] FIG. 2 illustrates a block diagram representation of an exemplary computing system 102, such as those shown in FIG. 1, for managing behaviour-based personalized services for end users 116, in accordance with an embodiment of the present disclosure. Further, the computing system 102 may include one or more hardware processors 202, a memory 204 and a storage unit 206.
[0063] Further, the one or more hardware processors 202, the memory 204 and the storage unit 206 are communicatively coupled through a system bus 208 or any similar mechanism. Further, the memory 204 may include a plurality of modules 110 in the form of programmable instructions executable by the one or more hardware processors 202.
[0064] Further, in an embodiment of the present disclosure, the one or more hardware processors 202, as used herein, means any type of computational circuit, such as, but not limited to, a microprocessor unit, microcontroller, complex instruction set computing microprocessor unit, reduced instruction set computing microprocessor unit, very long instruction word microprocessor unit, explicitly parallel instruction computing microprocessor unit, graphics processing unit, digital signal processing unit, or any other type of processing circuit. Further, the one or more hardware processors 202 may also include embedded controllers, such as generic or programmable logic devices or arrays, application specific integrated circuits, single-chip computers, and the like.
[0065] Further, the memory 204 may be non-transitory volatile memory and non-volatile memory. The memory 204 may be coupled for communication with the one or more hardware processors 202, such as being a computer-readable storage medium. Further, the one or more hardware processors 202 may execute machine-readable instructions and/or source code stored in the memory 204. A variety of machine-readable instructions may be stored in and accessed from the memory 204. Further, the memory 204 may include any suitable elements for storing data and machine-readable instructions, such as read only memory, random access memory, erasable programmable read only memory, electrically erasable programmable read only memory, a hard drive, a removable media drive for handling compact disks, digital video disks, diskettes, magnetic tape cartridges, memory cards, and the like. In the present embodiment, the memory 204 includes the plurality of modules 110 stored in the form of machine-readable instructions on any of the above-mentioned storage media and may be in communication with and executed by the one or more hardware processors 202.
[0066] Further, the storage unit 206 may be a cloud storage. The storage unit 206 may store the website data, properties, rules, and the like.
[0067] In an example embodiment, the processor 202 is configured to periodically obtain a multi-modal data associated with a user 116 from a plurality of data sources 108. The multi-modal data includes a behavioural data, an emotional data, a demographic data, and an ethnographic data. Behavioural data includes information derived from user interactions, engagement patterns, and decision-making tendencies across digital and physical environments. Examples of behavioural data may include, for example, browsing history, app usage frequency, time spent on specific content, purchase history, mouse clicks, search queries, chatbot interactions, and social media activity. Additionally, behavioural data may include, for example, implicit signals such as hesitation before making a selection, scrolling behaviour, and patterns of feature engagement within an application.
[0068] Emotional data refers to information that captures a user's emotional state, mood fluctuations, and sentiment responses based on their interactions with digital systems. Examples of emotional data may include, for example, sentiment analysis of text inputs, voice tone and pitch variations during speech-based interactions, facial expressions detected via image or video processing, and physiological indicators such as heart rate variability or stress levels recorded by wearable devices. Emotional data may also be inferred from engagement metrics, such as a user's reaction to personalized recommendations, frustration signals detected through abrupt session termination, or satisfaction measured through feedback ratings.
[0069] Demographic data includes quantifiable user characteristics that define their identity and socio-economic attributes. Examples of demographic data may include, for example, age, gender, income level, education background, marital status, occupation, and geographical location. These data points are often sourced from user-provided registration details, government records, survey responses, or publicly available databases. Demographic data plays a crucial role in segmenting users 116 into meaningful groups to tailor services, recommendations, and marketing strategies.
[0070] Ethnographic data encompasses cultural, linguistic, behavioural, and lifestyle attributes that provide deeper context into a user's social and environmental influences. Examples of ethnographic data may include, for example, language preferences, religious beliefs, cultural traditions, dietary habits, local customs, and community affiliations. This data may be collected through ethnographic studies, user interviews, observational research, or digital behaviour analysis within culturally specific applications. Ethnographic data is particularly useful for hyper-personalized services, as it enables customization based on cultural sensitivities, regional buying behaviours, and societal norms.
[0071] Further, the processor 202 is configured to determine user emotions and behavioural attributes associated with the user 116 based on the obtained multi-modal data using artificial-intelligence model. The behavioural attributes comprises a user interaction frequency, time spent on features, and content preferences. The user emotions comprises a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations. The behavioural attributes comprise at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies. The demographic and ethnographic factors comprise a location, a cultural background, an economic status, and language preferences of the user, and wherein the attitudinal interactions comprise analysing user decisions, preferences, and emotional responses during digital interactions.
[0072] In an example embodiment, the artificial intelligence model may include, for example, but not limited to, at least one of Natural Language Processing (NLP) models such as, for example BERT (Bidirectional Encoder Representations from Transformers) or GPT (Generative Pre-trained Transformer) for sentiment analysis and emotional inference from text-based interactions. Further, the artificial intelligence model may include, for example, but not limited to, Speech emotion recognition models such as wav2vec, Deep Speech, or LSTM-based (Long Short-Term Memory) recurrent neural networks for analysing tone, pitch, and vocal variations to detect emotional states. Also, the artificial intelligence model may include, for example, but not limited to, Computer vision models such as Open Face, AffectNet, VGGFace, or convolutional neural networks (CNNs) for facial expression recognition and emotion classification. Furthermore, the artificial intelligence model may include, for example, but not limited to, Physiological signal processing models such as DeepConvLSTM, BiLSTM, or transformer-based models for analysing heart rate variability, galvanic skin response (GSR), or EEG signals to determine stress, arousal, or cognitive load. Additionally, the artificial intelligence model may include, for example, but not limited to, User interaction-based models such as reinforcement learning models, Hidden Markov Models (HMMs), or behavioural clustering algorithms for detecting engagement patterns, decision-making tendencies, and implicit user preferences.
[0073] In an example embodiment, to determine the user emotions and the behavioural attributes associated with the user 116based on the obtained multi-modal data using the artificial-intelligence model, the processor 202 is configured to receive one or more user responses to a series of visual and interactive interfaces. The visual and interactive interfaces comprises generative artificial intelligence (Gen AI) models. The Generative Artificial Intelligence (Gen-AI) models may include, for example, but not limited to, AI-Powered Conversational Avatars, Adaptive UI Personalization Systems, Interactive AI-Generated Storytelling Environments, AI-Driven Emotion Recognition Interfaces, Generative AI-Powered Design Assistants, Real-Time Virtual Try-On Systems, Gamified Behavioural Profiling Interfaces, 3D Metaverse and Virtual World Generation, Dynamic Chatbots with Personalized Visual Feedback, AI-Generated Data Visualization Interfaces and the like.
[0074] The AI-Powered Conversational Avatars may refer to a virtual AI assistant that uses text-to-image and text-to-video generation to create real-time animated facial expressions and gestures based on user interactions. For example, D-ID's AI avatars that engage in personalized conversations by responding with dynamically generated visuals.
[0075] The Adaptive UI Personalization Systems may refer to a Gen-AI-driven dashboard that dynamically modifies colour schemes, layouts, and widgets based on the user's emotional state and behaviour. For example: Deep learning models (such as GANs or VAEs) generating real-time UI changes in a mobile banking app based on user stress levels. The Interactive AI-Generated Storytelling Environments may refer to a system where users 116 interact with an evolving story by selecting paths, and the Gen-AI model generates unique visual narratives in real-time. For example, Stable Diffusion or DALL¬?E 3 generating images or videos that adapt to user choices in an educational app or entertainment platform. The AI-Driven Emotion Recognition Interfaces may refer to a video conferencing tool that uses computer vision-based Gen-AI models to detect facial expressions and generate real-time animated avatars representing user emotions. For example, OpenAI's Whisper (speech-to-text) combined with GAN-based facial generation for enhanced digital interactions.
[0076] The Generative AI-Powered Design Assistant may refer to an interactive graphic design tool where users 116 provide a sketch, and the AI model generates detailed artwork, UI designs, or marketing materials. For example: Runway ML's AI tools for generating design elements dynamically based on user preferences.
[0077] The Real-Time Virtual Try-On Systems may refer to an AI-powered fashion assistant that uses GAN-based body tracking and image synthesis to generate virtual outfits based on user-selected styles and body measurements. For example: Zalando's AI try-on using StyleGAN to create realistic images of how clothing fits the user.
[0078] The Gamified Behavioural Profiling Interfaces may refer to a game-based personality assessment where a Gen-AI system generates unique scenarios in real time based on user reactions and decisions. For example, Reinforcement learning models generating adaptive challenges in an interactive mental health app.
[0079] The 3D Metaverse and Virtual World Generation may refer to a virtual environment generator where users 116 describe a scene, and a Gen-AI model creates an interactive 3D world. For example, NVIDIA's Omniverse using GAN-based scene generation to create immersive metaverse spaces dynamically.
[0080] The Dynamic Chatbots with Personalized Visual Feedback may refer to a chatbot that generates real-time animated avatars or emojis reflecting the conversation's sentiment using multimodal Gen-AI models. For example, ChatGPT integrated with Synthesis AI avatars for interactive customer service experiences.
[0081] The AI-Generated Data Visualization Interfaces may refer to a dashboard that uses Gen-AI to generate visual reports, graphs, and predictive analytics dynamically based on user queries. For example: ChatGPT Code Interpreter generating visual insights in finance or healthcare applications.
[0082] Further, the processor 202 is configured to identify macro-level behavioural patterns of the user 116 based on the received one or more user responses. The Macro-level behavioural patterns may refer to broad, high-level trends and recurring behaviours exhibited by users 116 over extended periods or across large populations. These patterns are typically derived from aggregated user data and provide insights into general user tendencies, preferences, and engagement trends rather than focusing on individual actions. Examples of macro-level behavioural patterns may include, for example, but not limited to, purchasing trends across demographic segments, peak usage times for digital platforms, general sentiment trends in customer feedback, and common navigation paths within an application or website. These patterns help in segmenting users 116 into broad categories, such as frequent buyers, passive users 116, or high-engagement customers, and are often used for large-scale personalization, targeted marketing, and system optimization. Macro-level behavioural analysis leverages machine learning models, statistical aggregation techniques, and clustering algorithms to detect trends across different contexts, such as social media engagement, financial spending habits, or content consumption behaviours. By understanding macro-level behavioural patterns, businesses and AI-driven systems may optimize recommendations, predict user needs on a broad scale, and enhance overall user experience.
[0083] Further, the processor 202 is configured to generate behavioural patterns of the user 116 based on the identified macro-level behavioural patterns. Behavioural patterns may refer to the recurring actions, decision-making tendencies, and interaction habits exhibited by a user 116 across different contexts and platforms. Generating behavioural patterns of the user 116 based on identified macro-level behavioural patterns involves refining broad trends into more personalized, detailed insights that reflect an individual’s unique preferences and engagement tendencies. This process begins with analysing macro-level behavioural patterns, which represent aggregated trends observed across a larger population segment, such as peak usage times, general sentiment trends, and common interaction flows.
[0084] To generate individual behavioural patterns, artificial intelligence (AI) models, such as for example, clustering algorithms, sequence modelling techniques (e.g., Hidden Markov Models or LSTMs), and reinforcement learning frameworks, may be used to process user-specific data within the context of these macro trends. Multi-modal data sources such as browsing activity, transaction history, engagement duration, social interactions, and even biometric responses contribute to refining the user’s behavioural profile. The AI system 102 detects deviations, consistencies, and evolving preferences by continuously comparing individual actions against macro-level benchmarks.
[0085] For instance, if a macro-level pattern indicates that most users 116 engage with a financial planning app during the first week of the month, but a specific user 116 consistently interacts with investment features during the last week, the computing system 102 generates a personalized behavioural pattern that highlights the user’s preferred financial planning schedule. Similarly, if macro-level data suggests that users 116 typically abandon a shopping cart after three days, but an individual exhibits delayed but eventual purchasing behaviour, the computing system 102 adapts its engagement strategy, such as delaying promotional reminders or offering personalized discounts at strategic intervals.
[0086] Once generated, behavioural patterns are stored within a user persona model and updated dynamically as new interactions are recorded. These patterns drive personalization engines, predictive analytics, and AI-driven decision-making processes, allowing systems to anticipate user needs, optimize recommendations, and deliver contextually relevant experiences in real-time.
[0087] Further, the processor 202 is configured to determine the user emotions and the behavioural attributes associated with the user 116 based on the generated behavioural patterns of the user. In an example embodiment, determining user emotions and behavioural attributes based on generated behavioural patterns involves analysing a user’s recurring actions, interaction tendencies, and engagement signals to infer emotional states and personality traits. This process relies on machine learning models, natural language processing (NLP), computer vision, and physiological data analysis to extract meaningful insights from behavioural patterns observed across various digital interactions. The first step in this determination process is behavioural pattern analysis, where AI models examine frequency, duration, intensity, and context of user interactions. For example, if a user 116 frequently engages with stress-relief content, rapidly switches between different applications, or exhibits irregular interaction pauses, the computing system 102 may infer restlessness or anxiety as a behavioural attribute. Similarly, if a user 116 spends prolonged time on specific features, repeatedly revisits decision-making interfaces, or displays hesitation in completing actions, this may indicate indecisiveness or cautious behaviour. By analysing such patterns, AI-driven systems may classify users 116 into engagement profiles, such as passive, exploratory, goal-oriented, or highly engaged users 116. To determine user emotions, multi-modal AI models integrate data from text, voice, facial expressions, and physiological signals. Natural Language Processing (NLP) models such as BERT or GPT-based sentiment analysis tools analyse text-based interactions, chat messages, and feedback responses to detect sentiment and emotional tone (e.g., positive, neutral, or negative). For voice-based interactions, speech emotion recognition models such as wav2vec or Deep Speech analyse pitch, tone, intonation, and speech rhythm to classify emotions such as excitement, frustration, or sadness. Computer vision models like Open Face, Affect Net, or CNN-based facial recognition systems analyse facial expressions captured via images or video streams to detect emotions such as joy, anger, or confusion. Additionally, physiological data models such as DeepConvLSTM or BiLSTM process biometric indicators like heart rate variability, skin conductance (GSR), and pupil dilation to infer emotional states such as stress, relaxation, or cognitive overload.
[0088] Further, the processor 202 is configured to generate a user persona model based on the determined user emotions and the behavioural attributes associated with the user. The user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user.
[0089] In an example embodiment, to generate the user persona model based on the determined user emotions and the behavioural attributes associated with the user, the processor 202 is configured to determine the user emotions based on a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations. Sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variation may refer to a process of evaluating a user’s emotional state and psychological tendencies based on multiple data inputs. This analysis combines natural language processing (NLP), computer vision, speech recognition, and physiological signal processing to derive insights into how a user feels and how their emotional state evolves over time.
[0090] Sentiment analysis of user interactions involves analysing text-based communications, such as chat messages, emails, product reviews, or social media comments, to determine the user’s emotional tone. NLP models such as for example, BERT, GPT, or VADER (Valence Aware Dictionary and sentiment Reasoner) classify text into categories such as positive, neutral, or negative sentiment. For example, if a customer support chatbot receives a message stating, “I’m really disappointed with this service,” the model detects negative sentiment and may escalate the issue for resolution.
[0091] Facial expression analysis uses computer vision models such as, for example, AffectNet, OpenFace, or convolutional neural networks (CNNs) to detect microexpressions and classify emotions such as happiness, sadness, anger, or surprise. For example, in a virtual meeting, an AI-powered system may analyse a participant’s facial cues to detect frustration when they furrow their brows or tighten their lips, enabling automated adjustments such as suggesting a break or simplifying content delivery.
[0092] Speech pattern analysis evaluates tone, pitch, tempo, and volume using speech emotion recognition models like wav2vec, Deep Speech, or LSTMs. A system analysing a call centre conversation, for instance, may detect rising pitch and increased speaking speed as signs of frustration, prompting an AI-driven assistant to offer de-escalation strategies. Similarly, a voice assistant may detect calm or enthusiastic speech tones and adjust its responses accordingly.
[0093] Biometric signal analysis may involve tracking heart rate variability, galvanic skin response (GSR), and pupil dilation to infer emotional states. Deep learning models such as, DeepConvLSTM or BiLSTM process these signals to detect stress, anxiety, or relaxation levels. For example, a fitness tracker may analyse an elevated heart rate and skin conductance while a user 116 reads an email, suggesting that the email might have triggered stress.
[0094] Emotional stability and mood variation analysis examines long-term emotional trends and fluctuations to assess a user's psychological patterns. AI models track daily sentiment scores, physiological signals, and interaction history to determine emotional consistency or instability. For instance, if a mental wellness app detects frequent mood swings through inconsistent sentiment patterns in journal entries and increased physiological stress markers, it may recommend meditation exercises or alert the user’s support network.
[0095] Further, the processor 202 is configured to identify the behavioural attributes comprising at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies. Furthermore, the processor 202 is configured to establish a correlation between the determined user emotions and the behavioural attributes using a machine learning model trained on a real-time user data and a historical user data. Furthermore, the processor 202 is configured to periodically update the user persona model based on real-time user interactions to indicate changes in the user emotions and the behavioural attributes. Additionally, the processor 202 is configured to structure the user persona model using a nonlinear profiling model to represent the user emotions and the behavioural attributes. Finally, the processor 202 is configured to generate a graphical representation of the structured user personal model using a weighted graph structure. The graphical representation comprises one or more nodes corresponding to emotional states and the behavioural attributes. The graphical representation comprises one or more edges corresponding to a correlation strength value between the one or more nodes. Some examples of graphical representations, may include, knowledge graphs, ontologies, and correlation networks. A knowledge graph may be constructed where nodes represent user emotional states (e.g., happy, anxious, frustrated) and behavioural attributes (e.g., purchase frequency, response time, engagement level). The edges between nodes are weighted based on observed correlations in user behaviour. For example, a user with a high engagement level might be highly correlated with a positive emotional state, while low engagement could be linked to frustration or dissatisfaction. The graph structure may include nodes: "Happy," "Frustrated," "Impulse Buying," "Session Duration," "Product Affinity", and edges: (Happy ‚Ui Impulse Buying, weight = 0.8), (Frustrated ‚Ui Session Duration, weight = -0.6). Further, an ontology may define relationships between emotions, behavioural traits, and interaction types in a hierarchical manner. The graph may use class-subclass relationships to categorize different user attributes. For example, a root Node: "User Behaviour", Sub-Nodes: "Emotional Responses," "Interaction Preferences", "Emotional Responses" ‚Ui ("Excited," "Neutral," "Bored"), "Interaction Preferences" ‚Ui ("Prefers Text Chat," "Prefers Voice Chat"). This ontology may help personalize chatbot interactions based on the user's past emotional responses. A correlation network graph may be generated where nodes represent behavioural metrics (click-through rate, purchase history, and the like), and edges represent statistical correlations between them. Stronger relationships are indicated with heavier-weighted edges. For example, nodes: "Page Visit Duration," "Click-Through Rate," "Customer Satisfaction Score" and Edges: "Page Visit Duration ‚Ui Click-Through Rate" (Weight = 0.75, indicating high correlation) The model may predict high-value customers by analysing the strongest correlations between behavioural attributes.
[0096] In yet another example, a temporal weighted graph may be used to model how user emotions and behaviours evolve over time. The edge weights dynamically adjust based on recent user interactions. For example, if a user starts with a neutral emotional state but, after a series of failed transactions, shifts to frustration, the graph may dynamically update the correlation strength between "Failed Transactions" and "Frustration." These graphical models provide an intuitive and structured approach to understanding user behaviours, enabling AI-driven personalization, predictive analytics, and real-time adaptation of services based on user emotions and preferences.
[0097] In an example embodiment, an e-commerce recommendation system may track how often a user 116 visits a product page (interaction frequency), how long they spend viewing items (engagement levels), how quickly they respond to purchase prompts (response time), their most frequently selected categories (preferences), and whether they tend to abandon carts before checkout (decision-making tendencies). This data is collected in real-time and stored in a historical database for behavioural analysis.
[0098] Further, the processor 202 may establish a correlation between the determined user emotions and behavioural attributes using a machine learning model trained on real-time and historical user data. The machine learning model may include a deep neural network (DNN), a decision tree-based model (such as XGBoost), or a recurrent neural network (RNN) for sequential behaviour prediction. For example, if a user 116 frequently abandons online forms, an AI system may analyse historical frustration signals (e.g., rapid typing, erratic mouse movements, or increased response time) and real-time emotion recognition (e.g., detected stress in voice commands) to infer decision fatigue. The model establishes a correlation, indicating that users 116 with similar emotional responses tend to disengage at certain steps, allowing proactive intervention such as a simplified form layout or AI-assisted guidance.
[0099] To ensure dynamic adaptability, the processor 202 may periodically update the user persona model based on real-time user interactions to reflect changes in user emotions and behavioural attributes. For instance, in a mental wellness application, if a user initially exhibits low engagement with meditation features but gradually increases usage over weeks, the processor 202 may update their persona model from a "low-engagement stress-prone user" to a "moderate-engagement wellness-adopting user". Similarly, if a streaming service detects shifts in content preferences, such as a user 116 moving from action movies to documentaries, the persona model is dynamically refined to enhance content recommendations.
[00100] Furthermore, the processor 202 may structure the user persona model using a nonlinear profiling model to represent the user’s emotions and behavioural attributes more accurately. Unlike traditional linear models that assume direct proportional relationships, a nonlinear profiling model captures complex, non-linear dependencies between emotional states and behaviours. For example, in a customer service chatbot, the computing system 102 recognizes that a high frequency of chatbot interactions combined with negative sentiment in text responses does not linearly imply dissatisfaction but rather indicates an urgent issue requiring escalation. This nonlinear profiling model ensures greater flexibility and precision in understanding diverse user behaviours.
[00101] Finally, the processor 202 generates a graphical representation of the structured user persona model using a weighted graph structure. This graph-based visualization comprises nodes representing emotional states and behavioural attributes and edges representing correlation strength values between them. For instance, in a personalized learning platform, the processor 202 may generate a graph where nodes represent engagement levels (high, medium, low), emotional states (frustrated, motivated, neutral), and study preferences (videos, quizzes, reading materials). The edges between nodes are weighted based on historical correlations, such as "Users 116 with high engagement and motivation tend to prefer quizzes over reading materials". If a student’s engagement level drops, the processor 202 may dynamically adjust recommendations by increasing interactive content to improve learning outcomes. The processor 202 then continuously refines the user persona model, enabling adaptive personalization, real-time sentiment-aware recommendations, and dynamic behavioural analysis across diverse applications.
[00102] In an example embodiment, the processor 202 is configured to compute a behavioural score for the user 116 based on the generated user persona model. The behavioural score corresponds to a unique value assigned to the user. To compute the behavioural score for the user 116 based on the generated user persona model, the processor 202 is configured to classify implicit and explicit behaviours of the users 116 based on external user data comprising visual inputs, heatmaps, chatbot interactions, and user engagement tracking. Further, the processor 202 is configured to extract demographic and ethnographic factors associated with the user 116 by performing attitudinal interactions with the user. Furthermore, the processor 202 is configured to analyse the behavioural attributes and the user emotions exhibited by the user 116 during the attitudinal interactions. Also, the processor 202 is configured to compute the behavioural score for the user 116 based on the analysed behavioural attributes and the user emotions.
[00103] In an example embodiment, the behavioural score indicates a user preferences, a user lifestyle, and an economic status. The Generative AI (Gen-AI) models may analyse both explicit and implicit behavioural data of the user. For example, the explicit behavioural data may include structured financial records such as, for example, credit scores, property purchase history, income levels, and bank statements. On the other hand, the implicit behavioural data is derived from social media activity, online browsing behaviour, interaction patterns, and visual content analysis. By integrating these data sources, a more holistic user profile may be created.
[00104] For example, an AI model may analyse images posted on a user's social media profile to extract insights about their purchasing behaviour and brand affiliations. For example, if a user posts a photo holding a high-end luxury bag, the processor 202 may apply image recognition techniques to identify the brand of the bag and classify it under a "luxury brand ownership" category. This information is then cross-referenced with explicit financial data, such as the user’s credit score, past shopping transactions, or bank statements, to determine their economic status and spending capacity.
[00105] Additionally, social media interactions and geotagged locations may provide valuable insights. If a user frequently shares holiday photos from exotic destinations, stays at five-star hotels, and engages with luxury car brands, these implicit behaviours may be mapped to financial indicators to refine the behavioural score. For example, a user who frequently visits high-end restaurants and shopping districts may be assigned a higher spending potential score, even if their credit score is average.
[00106] By leveraging multi-modal AI techniques, including Natural Language Processing (NLP) for text analysis, Computer Vision for image recognition, and Machine Learning models for predictive analytics, the processor 202 generates the behavioural score that represents a multi-dimensional behavioural profile of the user. This enables businesses, financial institutions, and personalized recommendation engines to tailor services, product offerings, and credit risk assessments based on the user’s actual lifestyle and behavioural patterns, rather than relying solely on traditional credit reports.
[00107] Thus, the integration of explicit and implicit behavioural data sources using Gen-AI creates a more accurate and dynamic representation of user preferences and financial behaviour, facilitating smarter decision-making for businesses and enhanced personalization for consumers.
[00108] In an example embodiment, to compute the behavioural score for the user 116 based on the generated user persona model, the processor 202 first classifies implicit and explicit behaviours by analysing external user data sources, including visual inputs, heatmaps, chatbot interactions, and user engagement tracking. Implicit behaviours, which are unconscious or passive user actions, may be inferred from mouse movement patterns, session duration, hesitation before clicking, or gaze fixation on a product or interface element. Explicit behaviours, which involve conscious decision-making, include for example, form submissions, purchase completions, direct feedback, and chat interactions. For example, in a real estate platform, if a user spends extended time hovering over high-end property listings (implicit behaviour) but applies filters for budget-friendly homes (explicit behaviour), the processor 102 classifies them as a "luxury home researcher" who is still price-sensitive. This classification helps adjust recommendations accordingly.
[00109] Further, the processor 202 extracts demographic and ethnographic factors associated with the user 116 by performing attitudinal interactions, which involve direct or indirect engagement activities that reveal user preferences and decision-making styles. These interactions may include survey responses, adaptive questioning in chatbots, user-selected content categories, or responses to gamified preference assessments. For example, in a travel booking application, the processor 202 may present users 116 with an image-based preference test, asking them to choose between beach vacations, cultural excursions, or adventure sports. By analysing selections and response times, the processor 202 extracts demographic insights (e.g., preferred destinations based on location and past bookings) and ethnographic factors (e.g., cultural influences on travel preferences, such as a preference for historical sites over nightlife experiences).
[00110] After collecting these inputs, the processor 202 analyses the behavioural attributes and user emotions exhibited during attitudinal interactions. The processor 202 applies natural language processing (NLP) models, sentiment analysis tools, and computer vision techniques to understand how a user 116 reacts to different options, engagement prompts, or interactions. For example, in an AI-powered education platform, if a student frequently skips video content but engages more with quizzes, and sentiment analysis of chatbot interactions detects frustration when watching tutorials, the processor 202 infers that the user 116 prefers interactive learning over passive video consumption. Similarly, if heatmap analysis on an e-commerce platform reveals prolonged focus on "customer reviews" rather than "product descriptions," the processor 202 detects that the user 116 prioritizes peer feedback before making a purchase decision.
[00111] Finally, the processor 202 computes the behavioural score for the user 116 based on the analysed behavioural attributes and emotions. The behavioural score is assigned a numerical value that quantifies engagement, interaction tendencies, and inferred preferences, which may be dynamically updated based on real-time user interactions. For example, in a personal finance management app, if a user 116 frequently revisits investment options but never makes a transaction, their behavioural score may indicate "high interest, low action." However, if the user 116 begins exploring risk assessment tools or consulting financial advisors through the app, the score is updated to "investment-ready", triggering tailored recommendations such as personalized investment strategies or promotional offers on advisory services.
[00112] By systematically classifying behaviours, extracting demographic and ethnographic insights, analysing attitudinal interactions, and quantifying behavioural tendencies, the computed behavioural score enables hyper-personalized recommendations, adaptive content delivery, and dynamic user experience optimization across various applications.
[00113] In an example embodiment, in an AI-powered e-commerce platform, the processor 202 first collects and analyses multi-modal user data, including explicit behaviours (such as product searches, cart additions, purchases, and review submissions) and implicit behaviours (such as dwell time on product pages, scrolling patterns, and hesitation before clicking "Buy Now").
[00114] To compute the behavioural score, the AI model classifies and quantifies user interactions using machine learning techniques such as , for example, reinforcement learning, clustering algorithms, and deep neural networks. For example, a recurrent neural network (RNN) or a Transformer-based sequence model analyses a user’s browsing history to detect recurring shopping patterns. If a user 116 frequently browses high-end electronics but does not make purchases, the AI model assigns a high interest, low conversion score, indicating the user 116 is in the consideration phase. Conversely, if a user 116 frequently adds items to the cart and completes checkouts quickly, the AI model assigns a high engagement, high purchase intent score, signalling a conversion-ready user.
[00115] Additionally, sentiment analysis and NLP models analyse chatbot interactions, product reviews, and customer support queries to determine user sentiment and refine the behavioural score. If a user 116 asks a chatbot "When will this item be back in stock?" or leaves a review stating "I love this brand, but the last product did not meet my expectations", the AI model adjusts the behavioural score to reflect high brand affinity but moderate satisfaction, triggering targeted discounts or loyalty rewards to retain engagement.
[00116] The AI model also integrates computer vision models and heatmap analysis to evaluate user interactions with product images and UI elements. For instance, if a user 116 spends significant time zooming in on product images, reading detailed descriptions, and watching product videos, the AI model interprets this as a high-intent behaviour, increasing the behavioural score. However, if the user 116 frequently bounces between similar product pages without taking action, the AI model assigns a comparative shopper score, suggesting they need additional incentives, such as price-drop alerts or side-by-side product comparisons.
[00117] Biometric and contextual data further enhance behavioural scoring. If a mobile shopping app integrates with a user’s smartwatch, AI models may analyse heart rate variations during checkout, detecting stress or hesitation in high-value purchases. A sudden increase in heart rate before checkout may indicate purchase anxiety, leading the processor 202 to suggest flexible payment options like "Buy Now, Pay Later" or a one-time discount to encourage purchase completion.
[00118] The final behavioural score is computed dynamically using a weighted aggregation model, where different behavioural attributes (such as engagement frequency, purchase patterns, sentiment, and contextual factors) are weighted based on predictive purchase likelihood. If a user 116 with a high engagement but low conversion score suddenly starts interacting with flash sale notifications, the AI model recalculates the behavioural score in real time, shifting the user 116 into a "ready-to-purchase" category and triggering personalized promotions at the right moment.
[00119] Ultimately, the computed behavioural score is used for hyper-personalized recommendations and marketing automation. A user 116 with a high behavioural score and strong brand loyalty may receive early access to limited-edition products, exclusive offers, or tailored bundle deals. In contrast, a user 116 with a declining behavioural score (indicating decreased interest) may be re-engaged with targeted email campaigns, AI-driven retargeting ads, or chatbot-driven personalized assistance to revive shopping interest.
[00120] By leveraging AI-based behavioural scoring, the e-commerce platform ensures higher conversion rates, improved customer satisfaction, and more effective marketing strategies by understanding and predicting user behaviour in real-time.
[00121] Further, the processor 202 is configured to continuously train the user persona model based on user reactions to recommendations, real-time interactions of the user 116 with the one or more applications, and user activity across the one or more applications;. Furthermore, the processor 202 is configured to continuously update user profiles and recommendations through a feedback loop based on the trained user persona model. The processor 202 is further configured to dynamically adjust the computed behavioural score based on the updated user profiles and the recommendations.
[00122] In an example embodiment, the processor 202 is configured to predict one or more probable interactions of the user 116 with one or more applications based on computed behavioural score using generative artificial intelligence (Gen-AI) models. To predict the one or more probable interactions of the user 116 with the one or more applications 108A-N based on the computed behavioural score using the generative artificial intelligence (Gen-AI) models, the processor 202 is configured to detect sequential interaction patterns of the user 116 by analysing the behavioural score using a transformer-based neural network model. The processor 202 is further configured to generate a set of probable user interactions with one or more applications based on the detected sequential interaction patterns, a prior engagement history, and inferred preferences. Furthermore, the processor 202 is configured to rank the set of probable user interactions using a probabilistic scoring model based on confidence levels derived from the detected sequential interaction patterns and real-time engagement metrics. Further, the processor 202 is configured to tune the ranked probable user interactions by incorporating multi-modal inputs comprising at least one of a text data, a voice data, a visual data, a biometric feedback, and an environmental context. Additionally, the processor 202 is configured to simulate the tuned probable user interactions in a virtual environment using synthetic user profiles to validate model accuracy and response variations. Moreover, the processor 202 is configured to predict the one or more probable interactions of the user 116 with the one or more applications based on results of simulation.
[00123] In an example embodiment, to predict one or more probable interactions of the user 116 with the one or more applications 108A-N based on the computed behavioural score using generative artificial intelligence (Gen-AI) models, the processor 202 first detects sequential interaction patterns by analysing the behavioural score using a transformer-based neural network model. Transformer models, may include such as, for example, but not limited to, BERT (Bidirectional Encoder Representations from Transformers) or GPT (Generative Pre-trained Transformer), analyse past user behaviours by identifying recurring patterns in engagement, decision-making sequences, and content preferences. For example, in an e-commerce platform, if a user 116 has a history of viewing high-end fashion products every weekend but only making purchases during seasonal sales, the processor 202 detects a sequential interaction pattern linked to price sensitivity and periodic engagement.
[00124] Once these interaction patterns are detected, the processor 202 generates a set of probable user interactions with one or more applications by considering the detected sequential interaction patterns, prior engagement history, and inferred preferences. For instance, in a video streaming platform, if a user 116 consistently watches documentaries at night and action movies on weekends, the processor 202 predicts probable future interactions such as watching a newly released crime documentary on a weekday or engaging with an interactive action movie trailer on a Saturday. Similarly, in a smart home automation system, if a user 116 adjusts thermostat settings and dims the lights every evening, the processor 202 may predict a probable interaction where the user 116 schedules ambient lighting adjustments before bedtime automatically.
[00125] Further, the processor 202 ranks the set of probable user interactions using a probabilistic scoring model, which evaluates confidence levels derived from real-time engagement metrics and previously detected sequential interaction patterns. The ranking process assigns higher scores to interactions that closely align with the user's historical engagement trends, inferred emotional state, and behavioural score fluctuations. For example, in a ride-hailing app, if a user 116 frequently requests a ride at 8 AM on weekdays, the processor 202 ranks the probability of the user 116 needing a ride the next morning higher than other interactions, triggering a pre-ride booking suggestion.
[00126] To refine accuracy, the processor 202 further tunes the ranked probable user interactions by incorporating multi-modal inputs, including text data, voice data, visual data, biometric feedback, and environmental context. In a personalized fitness application, if a user 116 verbally states, “I feel tired today”, the processor 202 interprets the voice data sentiment and biometric heart rate readings to adjust the predicted workout recommendations, shifting from high-intensity training to a light stretching session. Similarly, in a smart vehicle system, if visual data from an in-car camera detects drowsiness, biometric sensors register increased fatigue, and environmental data indicates nighttime driving, the processor 202 predicts a probable interaction where the user 116 activates lane assist or pulls over for rest recommendations.
[00127] After tuning the interactions, the processor 202 simulates the tuned probable user interactions within a virtual environment using synthetic user profiles to validate the model’s accuracy and response variations. The simulation phase involves testing AI-generated predictions across multiple user scenarios, such as how a recommendation engine reacts to different user preferences or how a chatbot interacts with various sentiment expressions. For example, in an AI-driven financial advisory platform the processor 202 runs simulations using synthetic profiles of risk-averse and high-risk investors to validate whether the AI correctly predicts investment interactions, such as stock purchases or savings recommendations, under different market conditions.
[00128] Finally, the processor 202 predicts the one or more probable interactions of the user 116 with the one or more applications based on the results of the simulation. If the simulation confirms high accuracy and alignment with real-world user behaviour, the processor 202 deploys the prediction model into live interactions, adjusting recommendations, service triggers, and automation processes accordingly. For example, in a voice-enabled smart assistant, if simulations indicate that users 116 prefer shorter responses during morning hours but detailed explanations in the evening, the processor 202 dynamically adapts its interaction style based on predicted user behaviour.
[00129] By following these steps, the processor 202 continuously refines user interaction predictions, ensuring hyper-personalized recommendations, real-time adaptability, and an intuitive, context-aware user experience across multiple applications.
[00130] Further, the processor 202 is configured to determine one or more modifications to be made to service configurations of the one or more applications 108A-N by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications. The one or more virtual applications correspond to computer simulated version of the one or more applications 108A-N. Furthermore, the processor 202 is configured to modify the service configurations of the one or more applications based on the determined one or more modifications. The service configurations comprises at least one of a network throughput, a response time, and inter-service links.
[00131] In an example embodiment, to determine modifications to service configurations for one or more applications 108A-N, the processor 202 first simulates the computed behavioural score and predicted future interactions onto one or more virtual applications. These virtual applications serve as computer-simulated versions of real-world applications, enabling AI-driven testing and evaluation of how user behaviour impacts system performance and service configurations before deployment.
[00132] For example, in a video streaming platform, the processor 202 simulates a user 116 with a high behavioural score for binge-watching, predicting future interactions such as frequent high-resolution streaming and fast-forwarding through previews. The virtual application then tests the impact of these behaviours on network bandwidth, buffering rates, and response time, allowing the processor to detect potential congestion issues and determine modifications such as adaptive bitrate streaming or preloading frequently watched content to optimize performance.
[00133] Similarly, in an e-commerce platform, if the processor 202 predicts a surge in high-intent shopping behaviour during a flash sale event, the virtual application simulates thousands of users 116 interacting simultaneously, assessing server load, checkout speed, and payment processing times. Based on the simulation results, the processor 202 determines whether to modify service configurations by increasing network throughput, load balancing requests, or enabling faster caching mechanisms to prevent slowdowns and ensure seamless transactions.
[00134] Once the modifications are determined, the processor 202 modifies the service configurations of the real-world applications based on the simulation results. These service configurations include network throughput, response time, and inter-service links. In a cloud gaming application, if simulation results show that users 116 with a low behavioural score for engagement tend to quit games due to lag, the processor 202 may modify network throughput by allocating additional server resources to reduce latency and improve responsiveness, ensuring a smoother gaming experience.
[00135] In a smart home automation system, if simulated user interactions predict frequent remote access to IoT devices, the processor 202 may modify inter-service links by optimizing API request handling between the mobile app and connected home devices. This modification ensures faster response times when adjusting smart lighting, security cameras, or temperature controls.
[00136] By iterating through simulations and real-world adjustments, the processor 202 continuously refines service configurations, ensuring that applications dynamically adapt to user behaviour in real time, enhancing performance, user experience, and system efficiency across multiple platforms.
[00137] In an example embodiment, to modify the service configurations of the one or more applications based on the determined one or more modifications, the processor 202 is configured to identify at least one application among the one or more applications intended to be accessed by the user 116 based on the determined user emotions and the behavioural attributes. Further, the processor 202 is configured to determine at least one service configuration of the identified at least one application relevant to the user 116 based on the computed behavioural score and the predicted future interactions. Furthermore, the processor 202 is configured to determine one or more modifications to be made to the determined at least one service configuration. Furthermore, the processor 202 is configured to modify the at least one service configuration based on the determined one or more modifications.
[00138] In an example embodiment, to modify the service configurations of one or more applications based on the determined modifications, the processor 202 first identifies at least one application among the available applications that the user 116 intends to access. This identification is based on the determined user emotions and behavioural attributes, which are derived from historical engagement patterns, real-time interactions, and inferred emotional states.
[00139] For example, in a smart home ecosystem, the processor 202 detects that the user 116 frequently adjusts lighting and thermostat settings in the evening, indicating a preference for personalized home automation. Based on this behavioural pattern and sentiment analysis from previous interactions (e.g., frustration detected when the app fails to adjust temperature settings instantly), the processor 202 identifies the smart home control application as the primary service the user 116 intends to access.
[00140] Once the relevant application is identified, the processor 202 determines at least one service configuration of the identified application that is relevant to the user. This determination is based on the computed behavioural score and the predicted future interactions. For instance, in a ride-hailing application, if the behavioural score suggests that the user 116 prioritizes faster ride options over cost savings, and the predicted interactions indicate a likelihood of booking multiple rides during peak hours, the processor 202 determines that network throughput and response time are the most relevant service configurations that require optimization to ensure seamless booking experiences.
[00141] Further, the processor 202 determines one or more modifications to be made to the identified service configuration. This step involves analysing real-time system performance data, user demand, and simulated load conditions to optimize resource allocation. In a cloud-based gaming platform, if the processor 202 predicts high engagement levels for competitive multiplayer gaming, the processor determines that the game’s latency settings must be adjusted, requiring a modification to increase network throughput and reduce lag spikes during gameplay.
[00142] Finally, the processor 202 modifies the at least one service configuration based on the determined modifications, ensuring that the application dynamically adapts to the user’s real-time needs. For example, in a financial trading application, if the user 116 exhibits high-frequency trading behaviours and a preference for real-time data feeds, the processor 202 increases API call priority, optimizes real-time data streaming, and reduces response time for trade execution. This modification ensures that market data refreshes more frequently, allowing the user 116 to make faster, more informed trading decisions.
[00143] By continuously identifying relevant applications, determining and modifying service configurations, and dynamically adapting system performance, the processor 202 ensures a seamless, personalized user experience across various applications while enhancing efficiency, responsiveness, and service reliability.
[00144] Further, the processor 202 is configured to output the computed behavioural score as a token to the one or more applications upon modifying the service configurations of the one or more applications 108A-N. In an example embodiment, the processor 202 is configured to encrypt the computed behavioural score using one or more cryptographic models and integrate the encrypted behavioural score as the token into the one or more applications using a blockchain network upon modifying the service configurations of the one or more applications. The token corresponds to a non-fungible token (NFT).
[00145] In an example embodiment, to output the computed behavioural score as a token to one or more applications (108A-N) after modifying their service configurations, the processor 202 first ensures that the behavioural score reflects the latest user interactions, engagement trends, and system optimizations. This score, which quantifies user engagement, preferences, and decision-making patterns, is then formatted into a machine-readable token for seamless integration across different applications.
[00146] Once the service configurations have been modified, the processor 202 encrypts the computed behavioural score using one or more cryptographic models to ensure data security, integrity, and user privacy. The encryption process employs advanced cryptographic techniques, such as, for example, but not limited to, Elliptic Curve Cryptography (ECC), AES (Advanced Encryption Standard), or SHA-256 hashing, to encode the behavioural score before it is shared with external applications. For example, in a personalized advertising platform, if a user’s behavioural score indicates a high interest in eco-friendly products, the processor 202 encrypts this score before embedding it into a digital token to be used for targeting relevant ads while ensuring that the user's raw data remains anonymized.
[00147] The processor 202 then integrates the encrypted behavioural score as a token into one or more applications using a blockchain network. The blockchain ensures that the tokenized behavioural score remains immutable, traceable, and verifiable across different applications. For example, in a cross-platform loyalty rewards system, a user's behavioural score calculated based on purchase frequency, engagement with promotions, and product preferences—is converted into an NFT (Non-Fungible Token). This NFT is then stored on a decentralized blockchain ledger, allowing multiple e-commerce platforms to authenticate and reward user engagement without directly exposing sensitive personal data.
[00148] The tokenized behavioural score (NFT) is then distributed to relevant applications, enabling real-time personalization, service enhancements, and cross-platform user profiling. For instance, in a metaverse-based retail environment, a user 116 with a high behavioural score for fashion-related purchases may have their NFT token automatically recognized by different virtual stores, unlocking exclusive discounts, early access to limited-edition items, or AI-driven styling recommendations. Similarly, in a blockchain-integrated healthcare system, a patient’s engagement score—derived from fitness app activity, diet tracking, and wellness program participation—may be tokenized and used to access personalized health services across different providers without sharing raw health data.
[00149] By encrypting and tokenizing the behavioural score using blockchain, the processor 202 ensures that user engagement data remains secure, transferable, and interoperable across multiple applications, while providing a decentralized, privacy-preserving approach to personalized services and digital asset management.
[00150] Further, the processor 202 is configured to generate a user token based on the computed behavioural score. The behavioural score is computed from the user persona model representing correlations between the user emotions and the behavioural attributes. The processor 202 is further configured to assign a cryptographic key pair to the generated user token using a cryptographic model. Furthermore, the processor 202 is configured to encrypt the user token using a blockchain-based encryption model upon assigning the cryptographic key pair. Also, the processor 202 is configured to transmit the encrypted user token to a distributed ledger network comprising a plurality of transactional nodes. Also, the processor 202 is configured to validate the encrypted user token using a consensus technique between the plurality of transactional nodes and record the encrypted user token onto the distributed ledger network as the non-fungible token (NFT). Furthermore, the processor 202 is configured to verify one or more token access requests received from the plurality of transactional nodes using a cryptographic authentication model.. The one or more token access requests comprises a cryptographic key for retrieving and decrypting the user token. Further, the processor 202 is configured to control access to the encrypted user token based on results of verification using smart contract-based permissions. The smart contract-based permissions comprise one or more conditions for token retrieval, decryption, and transfer.
[00151] The one or more conditions ensure that only authorized entities may access the behavioural data while adhering to GDPR, CCPA, and other data protection regulations. One fundamental condition is explicit user consent before any entity may access or retrieve the token. The computing system 102 requires the user's digital signature or multi-factor authentication (MFA) approval to verify their consent. For example, a retail platform requests access to a user's behavioural token to personalize product recommendations. The computing system 102 sends an approval request to the user's registered device, and the token is only retrieved upon explicit approval. Another conditions is for example, decryption of the user token is restricted based on predefined roles and access levels. Entities are categorized into different tiers, such as service providers, AI recommendation engines, or advertising networks, with each having limited access permissions. For example AI-based recommendation systems may have partial access to behavioural patterns but cannot decrypt sensitive personal attributes. Also, payment processors may only retrieve transactional behaviour insights without accessing emotional or preference data.
[00152] Another condition may be access to the token is granted only if the requesting entity complies with regulatory frameworks like GDPR (General Data Protection Regulation) and CCPA (California Consumer Privacy Act). The smart contract automatically checks compliance logs before allowing retrieval. For example, a European service provider requesting access to a European user's behavioural token must comply with GDPR regulations. If the computing system 102 detects non-compliance (e.g., missing Data Protection Impact Assessment (DPIA) records), the retrieval request is automatically denied.
[00153] Yet another condition may be that token retrieval or decryption may be limited to a specific time window or session-based access to prevent continuous or unauthorized tracking. For example, a video streaming platform requests access to a user's watching preferences. The behavioural token is only accessible for 24 hours, after which it requires re-authorization.
[00154] Still another condition may be Multi-Factor Authentication (MFA) for Token Transfer. Before transferring the behavioural token between services or applications, the computing system 102 requires multi-factor authentication (MFA) involving: Biometric authentication (e.g., fingerprint, facial recognition), One-time password (OTP) verification or Device authentication (e.g., blockchain wallet confirmation) or the like.
[00155] Yet another condition may be Zero-Knowledge Proof (ZKP)-Based Token Sharing. For enhanced privacy, Zero-Knowledge Proofs (ZKPs) allow an entity to verify behavioural patterns without revealing the actual data.
[00156] In an example embodiment, to generate a user token based on the computed behavioural score, the processor 202 first calculates the behavioural score by analysing the user persona model, which represents correlations between user emotions and behavioural attributes. This process involves assessing historical engagement trends, real-time user interactions, and inferred emotional responses. For example, in a personalized e-commerce platform, if a user 116 frequently engages with luxury fashion items but hesitates before making a purchase, the behavioural score may indicate high interest but cautious decision-making, leading to tailored instalment-based payment offers. The computed score is then formatted into a structured digital token for secure transmission and interoperability across different applications.
[00157] Once the token is generated, the processor 202 assigns a cryptographic key pair using a cryptographic model. This key pair consists of a public key, which may be used for verification, and a private key, which ensures secure authentication. For example, in a blockchain-based loyalty rewards system, the user's behavioural score—linked to purchase frequency and brand engagement—is encrypted with a cryptographic key pair, ensuring only authorized platforms may access and verify the score while keeping the user's personal details anonymized.
[00158] After generating the cryptographic key pair, the processor 202 encrypts the user token using a blockchain-based encryption model. This encryption ensures tamper resistance, privacy preservation, and secure cross-platform token transfer. For example, Elliptic Curve Cryptography (ECC), Advanced Encryption Standard (AES-256), or Zero-Knowledge Proofs (ZKP) may be used to encrypt the token. For instance, in a decentralized digital identity system, a user's engagement score—computed from social media activity, online learning participation, or gaming achievements—is encrypted on a blockchain-based self-sovereign identity (SSI) network, allowing the user 116 to control their behavioural credentials without exposing raw data.
[00159] The processor 202 then transmits the encrypted user token to a distributed ledger network, which comprises a plurality of transactional nodes responsible for processing, validating, and maintaining a decentralized record of the token. For example, in a blockchain-integrated healthcare system, a behavioural engagement score—computed from physical activity, wellness program adherence, and diet tracking—is transmitted to a health data ledger, allowing insurance providers and wellness partners to verify user engagement securely.
[00160] Once transmitted, the processor 202 validates the encrypted user token using a consensus technique between the plurality of transactional nodes. This ensures the authenticity and integrity of the stored token before it is permanently recorded onto the blockchain. The consensus technique may include , for example, but not limited to, Proof of Stake (PoS), Practical Byzantine Fault Tolerance (PBFT), or Proof of Authority (PoA). For example, in a digital financial credit scoring system, nodes validate a user’s behavioural token computed from spending habits and transaction history before allowing the score to be used for loan eligibility assessments or credit limit adjustments.
[00161] After validation, the processor 202 records the encrypted user token onto the distributed ledger network as a non-fungible token (NFT). This NFT-based behavioural token acts as a secure, verifiable digital identity representing the user’s behavioural tendencies. For example, in a metaverse-based economy, an NFT tokenized behavioural score derived from in-game achievements, trading patterns, and social interactions may be used across different virtual worlds to unlock exclusive content or personalized AI-driven recommendations.
[00162] To maintain security, the processor 202 verifies one or more token access requests received from the plurality of transactional nodes using a cryptographic authentication model. The access requests include a cryptographic key, which must be validated before retrieving and decrypting the user token. For example, in a privacy-focused advertising network, advertisers requesting access to a behavioural score for ad targeting must present a valid cryptographic key, ensuring that only verified partners may retrieve user insights without exposing sensitive personal data.
[00163] Finally, the processor 202 controls access to the encrypted user token based on smart contract-based permissions, ensuring that retrieval, decryption, and transfer operations follow predefined security conditions. These smart contract-based permissions enforce automated compliance with privacy regulations (e.g., GDPR, CCPA) and allow users 116 to define how and when their behavioural data may be used. For instance, in a decentralized finance (DeFi) lending platform, the behavioural token representing a user's history of responsible borrowing and repayment—may only be accessed by lenders who meet predefined trust criteria established in the smart contract, ensuring data integrity and secure transactions.
[00164] By following these steps, the processor 202 ensures secure, verifiable, and privacy-preserving behavioural score management using blockchain-based tokenization, enabling seamless cross-platform integration, decentralized identity verification, and AI-driven hyper-personalization.
[00165] Further, the processor 202 is further configured to generate one or more large language model (LLM) prompts based on the behavioural score and predicted user interactions to tune a user engagement level through a conversational AI. Furthermore, the processor 202 is further configured to simulate the predicted user interactions within a virtual environment using the Gen-AI model and the generated one or more LLM prompts to assess response patterns and tune interaction predictions prior to deployment. Additionally, the processor 202 is configured to generate one or more LLM responses for the generated one or more LLM prompts using the Gen-AI model based on outcomes of the simulated user interactions. The one or more LLM responses comprise one or more recommendations on user interactions, service configurations, user behavioural patterns, the behavioural attributes and the user emotions.
[00166] In an example embodiment, to tune user engagement levels through conversational AI, the processor 202 first generates one or more Large Language Model (LLM) prompts based on the behavioural score and predicted user interactions. These LLM prompts are structured to personalize AI-driven conversations and enhance user engagement by dynamically adapting to the user's emotional state, preferences, and decision-making tendencies.
[00167] For example, in an AI-powered customer service chatbot for an e-commerce platform, if a behavioural score indicates hesitation in completing purchases and the predicted interactions suggest frequent cart abandonment, the processor 202 generates an LLM prompt such as:
[00168] "I noticed you've been exploring some great deals today! Would you like help comparing features or setting up a price alert?"
[00169] This prompt aligns with the user's behavioural tendencies, encouraging engagement without being overly intrusive.
[00170] Once the LLM prompts are generated, the processor 202 simulates the predicted user interactions within a virtual environment using the Gen-AI model to assess response patterns and tune interaction predictions prior to deployment. The virtual environment is a controlled AI-driven simulation space where the computing system 102 evaluates how users 116 might respond to different prompts and interactions.
[00171] For instance, in a personalized fitness coaching app, if an AI assistant suggests workout plans based on prior engagement patterns, the processor 202 simulates multiple user personas (e.g., a highly motivated user vs. a user who frequently skips workouts). If the simulation reveals that users 116 disengage when offered complex fitness routines, the processor 202 re-tunes its recommendations to include simplified, beginner-friendly options before deployment.
[00172] After analysing simulated interactions, the processor 202 generates one or more LLM responses for the previously generated LLM prompts, using the Gen-AI model based on outcomes of the simulation. These responses are tailored to maximize engagement, improve user experience, and align AI recommendations with real-world behaviour.
[00173] For example, in a smart home automation system, if the processor 202 detects that a user 116 often adjusts thermostat settings manually despite having an automation feature enabled, the processor 202 reassesses past interactions and generates a refined LLM response:
[00174] "Would you like me to adjust the thermostat automatically based on your preferred schedule, so you don't have to set it manually every time?"
[00175] This refined response is more proactive and aligned with the user's behaviour, encouraging adoption of the automation feature.
[00176] The one or more LLM responses comprise recommendations on user interactions, service configurations, behavioural patterns, behavioural attributes, and user emotions. These recommendations help refine AI-driven decision-making across various applications.
[00177] For instance, in a financial planning app, if an AI assistant detects hesitation in investment decisions, it may suggest: "Based on your past choices, would you like to explore lower-risk investment options that align with your financial goals?". Similarly, in a ride-hailing service, if the processor 202 recognizes a preference for quiet rides, it may pre-emptively offer: "Would you like me to set your next ride to 'Do Not Disturb' mode?". Furthermore, in a mental wellness app, if behavioural analysis suggests fluctuating mood patterns, the processor 202 may refine its approach by recommending: "You've been completing your mindfulness exercises consistently. Would you like to try guided meditation to help with focus?". By continuously generating, simulating, and refining LLM prompts and responses, the processor ensures that AI-driven conversational interfaces remain adaptive, user-centric, and engagement-driven. This process allows for real-time personalization, improving user experience and optimizing service configurations across multiple applications.
[00178] Further, the processor 202 is configured to validate the predicted one or more probable interactions and the computed behavioural score using historical interaction logs and artificial intelligence model and generate one or more recommendations for modifying the service configurations based on results of validation.
[00179] In an example embodiment, to validate the predicted probable interactions and the computed behavioural score, the processor 202 first analyses historical interaction logs using an artificial intelligence (AI) model to assess how closely the predictions align with past user behaviours. These historical interaction logs include previous user actions, engagement patterns, decision-making tendencies, and feedback responses, allowing the processor 202 to compare predictions against actual past outcomes for accuracy.
[00180] For example, in an e-commerce platform, if the processor 202 predicts that a user 116 will purchase home fitness equipment based on their current behavioural score and browsing patterns, the processor 202 validates this prediction by examining historical data on similar users 116 who exhibited comparable browsing behaviour. If historical logs reveal that users 116 with similar engagement trends ultimately purchased accessories (such as yoga mats or resistance bands) before committing to expensive equipment, the processor 202 adjusts the probability of the predicted purchase behaviour accordingly.
[00181] Once validation is complete, the processor 202 generates one or more recommendations for modifying the service configurations based on the results of the validation process. This step ensures that service adjustments align with actual user behaviour rather than purely AI-generated predictions.
[00182] For example, in a ride-hailing application, if the processor 202 predicts that a user 116 will likely request a ride at a specific time based on their routine engagement patterns, the processor 202 validates this prediction using historical ride request logs. If historical data reveals that similar predictions led to inaccurate ride recommendations (e.g., user behaviour changes on weekends), the processor 202 modifies the ride availability model to dynamically adjust based on day-of-the-week behavioural variations.
[00183] Similarly, in a cloud gaming platform, if the processor 202 predicts that a user 116 will engage in extended multiplayer sessions due to a high behavioural score for competitive gaming, the processor 202 validates this prediction by checking past game session logs. If the historical data confirms that longer gaming sessions correlate with server congestion during peak hours, the processor 202 generates a recommendation to modify service configurations, such as pre-emptively scaling up server resources during anticipated high-demand periods.
[00184] The processor 202 also incorporates real-time validation feedback to refine AI-driven recommendations. In a personalized content streaming service, if the processor 202 predicts that a user 116 will engage with new movie releases based on their behavioural score, the processor 202 cross-checks historical watch patterns and real-time engagement metrics. If validation indicates that users 116 with similar engagement profiles are more likely to rewatch familiar genres instead of exploring new ones, the processor 202 modifies its recommendation strategy to prioritize personalized suggestions based on genre familiarity rather than trending releases.
[00185] By validating AI-driven predictions against historical interaction logs and modifying service configurations based on validation results, the processor 202 ensures that recommendations are dynamically optimized, reducing false predictions, improving personalization accuracy, and enhancing user 116 satisfaction across various applications.
[00186] FIG. 3 illustrates an example block diagram of a computing system 102 such as those shown in FIG. 1, capable of managing behaviour-based personalized services for end users 116, in accordance with an embodiment of the present disclosure.
[00187] Referring to FIG. 3, the computing system 102 may comprise one or more processor(s) 202 that may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices that process data based on operational instructions. Among other capabilities, the one or more processor(s) 202 may be configured to fetch and execute computer-readable instructions stored in a memory of the computing system 102. The memory 204 may be configured to store one or more computer-readable instructions or routines in a non-transitory computer readable storage medium, which may be fetched and executed to create or share data packets over a network service. The memory 204 may comprise any non-transitory storage device including, for example, volatile memory such as random-access memory (RAM), or non-volatile memory such as erasable programmable read only memory (EPROM), flash memory, and the like.
[00188] In an embodiment, the computing system 102 may include an interface(s) (not shown). The interface(s) may comprise a variety of interfaces, for example, interfaces for data input and output (I/O) devices, storage devices, and the like. The interface(s) may also provide a communication pathway for one or more components of the computing system 102.
[00189] In an embodiment of the present disclosure, the processor 202 may be implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the processor 202. In examples described herein, such combinations of hardware and programming may be implemented in several different ways. For example, the programming for the processor 202 may be processor-executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the processor 202 may comprise a processing resource (for example, one or more processors), to execute such instructions. In the present examples, the machine-readable storage medium may store instructions that, when executed by the processing resource, implement the processor 202. In such examples, the computing system 102 may comprise the machine-readable storage medium storing the instructions and the processing resource to execute the instructions, or the machine-readable storage medium may be separate but accessible to the computing system 102 and the processing resource. In other examples, the processor 202 may be implemented by electronic circuitry.
[00190] In an embodiment, the computing system 102 may be configured to provide a framework to learn and apply behavioural identifiers for the one or more end users 116. The framework may utilize the non-linear profiling model representing diverse human emotion and behavioural attributes. The outcome of the framework may drive identification and micro-segmentation of the one or more end users 116 through a combination of behavioural (e.g., both implicit and explicit), attributional interactions as well as demographic and ethnographic factors. Further, the computing system 102 may compute the behavioural identifier for micro-segmentation of targeted audience. In an aspect of the present disclosure, one or more end users 116 (shown in FIG. 1B) may become active by providing consent to data acquisition. The processor 202 may acquire and anonymise the data acquired from the user 116 to protect privacy and identity of the one or more end users 116. The computing system 102 may be provided with a behaviour pre-profiler module 303 to leverage the anonymized processed data to classify macro user segmentation based on implicit behavioural attributes. The one or more end users 116 may go through a series of visual and interactive interfaces, where responses of the one or more end users 116 may be used to identify macro level behavioural patterns of the one or more end users 116. In an example embodiment of the present disclosure, the one or more end users 116 may select an appropriate hat that may be close to their behaviour after which the behaviour pre-profiler module 303 may generate a pattern that later may be used by a multi-modal analyser module 305 to classify implicit and explicit behaviours. The computing system 102 may further be provided with a user interaction analyser module 304 that may be configured to capture user interaction and emotions through chatbot conversations by rating semantics. Further, the user interaction analyser module 304 may be configured to apply various analytics by capturing mouse clicks and heatmaps, filters, selection of properties, amenities and more importantly time spent on different Deep Insights. The computing system 102 may also be provided with a multi-modal analyser module 305 that may collect data from different sources, including, but are not limited to, visual inputs, heatmaps, and other forms of user interactions to form a comprehensive understanding of how users 116 engage with the computing system 102 or platform. The multi-modal analyser module 305 may apply inferential learning techniques to derive insights from the integrated data and make predictions or inferences based on patterns and correlations found in the collected data. The multi-modal analyser module 305 may be further configured to create a more detailed and accurate profile of the one or more end users 116, considering their behaviours across various modalities. As the one or more end users 116 continue to engage with the computing system 102, the multi-modal analyser module 305 may allow for adjustments to the user persona over time which is essential for capturing evolving preferences and behaviours. The multi-modal analyser module 305 may contribute to automatic morphing of user representations symbolized by "types of hats", and this morphing may reflect changes in user personas based on observed behaviours and interactions. In this regard, the multi-modal analyser module 305 ensures that user privacy is maintained by operating within the bounds of user consent and adhering to privacy and data security regulations.
[00191] In an embodiment of the present disclosure, the computing system 102 may be provided with a behaviour scoring module 306 that may be configured to establish certain behavioural metrics or criteria that are relevant to the personalized services to be provided to the one or more end users 116. The metrics may include factors like frequency of interaction, time spent on certain features, and the types of content accessed. Further, the behaviour scoring module 306 may apply scoring algorithms to assess and quantify user behaviours based on the defined metrics. The algorithms may be applied assign numerical scores to different aspects of user activity, reflecting the significance or relevance of each behaviour. The behaviour scoring module 306 may be configured to dynamically modify the assigned numerical scored over time based on evolving user interactions by considering recent behaviours more heavily and ensuring that the personalized services align with current preferences of the one or more end users 116. The scores generated may contribute to personalization efforts. Higher scores for specific behaviours may indicate a stronger affinity or interest, influencing the computing system 102 to prioritize recommendations aligned with those behaviours. Further, the behaviour scoring module 306 may analyse the scored data to identify patterns and trends in user behaviour which may help to uncover insights into user preferences and allow the computing system 102 to make informed predictions about future interactions. The behaviour scoring module 306 may be part of a feedback loop, where the scores influence the recommendations presented to the one or more end users 116. User reactions and responses to recommendations may further modify the numerical scores, creating a continuous improvement cycle.
[00192] In an embodiment of the present disclosure, the computing system 102 may be provided with an Artificial Intelligence (AI)/ Machine Learning (ML) based behaviour learning module 307 that may be configured to apply AI/ML techniques to understand and adapt to user behaviours. The AI/ML based behaviour learning module 307 may collect extensive data related to user interactions, preferences, and behaviours within the computing system 102 pertaining to user clicks, views, time spent on specific features, historical data, and any other relevant user activity to capture nuances of user behaviour. The AI/ML based behaviour learning module 307 may prepare a dataset for training the machine learning models. The dataset may comprise labelled examples such as, but not limited to, desired outcomes or behaviours of the one or more end users 116, allowing the model to learn behavioural patterns of the one or more end users 116. During training, the models may learn to recognize patterns in the data and establish connections between user behaviours and desired outcomes, such as preferences or engagement levels. Once trained, the models may make predictions about user behaviour. The AI/ML-based behaviour learning module 307 may analyse new data inputs and predict how the one or more end users 116 may be likely to interact with the computing system 102 or respond to specific content. The AI/ML based behaviour learning module 307 may further allow for dynamic adaptation by continuously updating the models based on new user data. The AI/ML based behaviour learning module 307 may also leverage the learned behaviour patterns to generate personalized recommendations and predicts what content or features would be most appealing to the one or more end users 116 based on their historical behaviours and preferences.
[00193] In an embodiment of the present disclosure, the computing system 102 may be provided with a behaviour identifier generator module 308 (also referred herein as behaviour chromatic identifier generator module 308) integrated with a cohort manager module 309 and a blockchain module 310 communicatively coupled to a ledger 311. The behaviour identifier generator module 308 of FIG. 3 may be configured to generate the unique behavioural ID or the behavioural token by computing the unique value and attaching the unique value with the unique behavioural ID. The unique behavioural ID or the behavioural token may be a Non-Fungible Token (NFT) that may be for ensuring transmissibility across various networks, multi-variant systems. The behaviour identifier generator module 308 may create profiles that represent various behavioural traits of the one or more end users 116 such as preferences, engagement patterns, decision-making styles, and other relevant aspects of user behaviour. The behaviour identifier generator module 308 may then assign the behavioural ID, often represented by colours, to different behavioural categories or personas. Each colour may correspond to a specific set of behavioural characteristics, creating a visual representation of user profiles.
[00194] In an embodiment of the present disclosure, the computing system 102 may also be provided with a service orchestrator module 312. The service orchestrator module 312 may manage the flow of data between different modules and processes involved in the computing system 102 which may include overseeing the transfer of, but is not limited to, user behaviour data, feedback loops, and information generated by the various recommendation components, and the like. The service orchestrator module 312 may facilitate static or real-time adaptation by ensuring that the computing system 102 may dynamically adjust to changes in user behaviour. The service orchestrator module 312 may manage the assignment of the unique behavioural ID to the one or more end users 116 based on their behavioural profiles. The service orchestrator module 312 may coordinate integration of the identifiers into the recommendation process. The service orchestrator module 312 may further incorporate feedback loops into the system and manage the collection and processing of user feedback on recommendations, facilitating continuous improvement in the accuracy and relevance of future recommendations. The service orchestrator module 312 may act as a central management component in the behaviour-based recommendation system, coordinating integration of various modules, managing data flow, and ensuring the system's adaptability, efficiency, and compliance with privacy and security standards.
[00195] In an example embodiment, FIG. 3 represents the computing system 102 that utilizes artificial intelligence (AI), machine learning (ML), blockchain, and behavioural profiling to generate, store, and manage behavioural tokens.
[00196] Specifically, the behaviour pre-profiler module 303 is configured to capture preliminary behavioural insights based on user history, preferences, and previous interactions. This behaviour pre-profiler module 303 helps in establishing a baseline behavioural pattern before deeper AI-based learning.
[00197] The user interaction analyser module 304 is configured to monitor and processes real-time user interactions across multiple platforms and services. Further, the user interaction analyser module 304 is configured to capture data such as click patterns, response times, engagement frequency, and decision-making tendencies.
[00198] The user multi-modal analyser module 305 is configured to collect and analyse multi-modal user data, including text-based sentiment analysis (e.g., chat interactions), voice-based emotion recognition, visual cues such as facial expressions, biometric signals (e.g., heart rate, stress levels) and the like. Further, the user multi-modal analyser module 305 provides a comprehensive emotional and behavioural profile of the user.
[00199] The behavioural scoring module 306 is configured to compute a behavioural score based on the analysed user interactions, emotions, and historical data. Further, behavioural scoring module 306 is configured to assign a unique numerical value to each user, representing their behavioural tendencies. Furthermore, the AI/ML-based behaviour learning module 307 uses artificial intelligence (AI) and machine learning (ML) to improve behavioural predictions and continuously refines behavioural models based on real-time and historical user data. Further, the AI/ML-based behaviour learning module 307 utilizes transformer-based models, neural networks, and generative AI to enhance behaviour analysis.
[00200] The behaviour chromatic identifier generator module 308 generates a chromatic behavioural identifier, which assigns a distinct color-coded or categorical label to the user based on their behavioural score. This identifier may be used in visual representations and clustering algorithms to group similar users 116. In an embodiment, the chromatic identifier is the user token.
[00201] The cohort manager module 309 categorizes users 116 into behavioural cohorts based on common patterns. Further, the cohort manager module 309 helps in grouping similar users 116 for personalized service delivery, targeted recommendations, and predictive analytics. Further, the cohort manager module 309 stores cohort data for further AI/ML-based analysis.
[00202] The blockchain module 310 ensures secure, immutable, and decentralized storage of user behavioural data. Further, the blockchain module 310 provides cryptographic protection to prevent unauthorized data access. The ledger 311 stores behavioural tokens along with service context, ensuring traceability and accountability. The behavioural tokens are securely encrypted and may be used for personalization, authentication, and secure transactions across multiple applications.
[00203] The service orchestrator module 312 manages interactions between AI models, behavioural modules, blockchain, and external applications. The service orchestrator module 312 ensures smooth integration of behavioural predictions into real-world applications and services.
[00204] Example Use Case: E-Commerce Personalization
[00205] A user 116 frequently browses fitness-related products and engages in wellness discussions online. The computing system 102 analyses behavioural data, assigns a high fitness engagement score, and categorizes the user 116 into a fitness-focused cohort. The e-commerce platform or the application 108 modifies service configurations to prioritize fitness-related product recommendations. A behavioural token is stored on the blockchain, ensuring secure, cross-platform personalization without compromising user 116 privacy.
[00206] FIG. 4 illustrates an example graphical user interface depicting a personality assessment or decision-making tool using different types of hats for personality assessment, in accordance with an embodiment of the present disclosure. In FIG. 4, a visual and interactive interface featuring different styles of hats such as hat 1, hat 2, hat 3, hat 4, hat 5, hat 6, hat 7, hat 8 is depicted. In an example, consider user A and user B. As part of a pre-profiling, user a chooses hat 1 implying a preference of luxury, opulence, and power of purchase, expecting higher service levels. User B chooses hat 2 implying risk averseness, cautious purchases, meticulousness of data and high degree of expectation in trustworthiness. When the user A enters the platform, the experience and service calibrate implicitly showing the items which have higher threshold of price, high-end brands, concierge services as well as curated recommendations. User B on the other hand is presented with items in moderate price thresholds however with copious data for the user to verify, including from pre-certified selections. Furthermore, the service offerings are adjusted based on the pre-profiling. User A receives limited stock and one of a kind of item recommendations whereas user B might be aligned with certified items or additional warranties due to the nature of their individual personas. As the one or more end users 116 perform tasks on the computing system 102, multimodal inputs including visual, heatmaps, decisions and more are used for inferential learning leveraging a training model which granularly calibrates the persona while also adjusting the hat type over time. In an example, if user A may repeatedly ignore or dismiss service offerings. In such a case upcoming items may not include such predictions/recommendations/service offerings. Alternatively, if the user A, despite choosing hat 1 behaves more conservatively around price choice, search and buying, the hat 1 may be morphed automatically to hat 3 which may signify a realignment of the user's persona post-profiled based on behaviour as well as inputs from the model.
[00207] FIG. 5 illustrates an example graphical user interface depicting chatbot conversations for managing behaviour-based personalized, in accordance with an embodiment of the present disclosure. FIG. 5. depicts a chatbot conversation screenshot of an exemplary interaction between the end user 116 and the computing system 102. The computing system 102 may capture user interaction and emotions through chatbot conversations by rating the semantics. The user interaction analyser module 304 associated with the computing system 102 may capture the user interaction and emotions through chatbot conversations by rating the semantics. Also, the user interaction analyser module 304 may gather various analytics by capturing mouse clicks and heatmaps, filters, selection of properties, amenities and more importantly time spent on different deep insights. Further, the user interaction analyser module 304 may generate a pattern for various analytics.
[00208] FIG. 6 illustrates an example blockchain network 600 depicting a process of transmission and validation of a user token for providing personalized services for end users 116, in accordance with an embodiment of the present disclosure. The FIG. 6 represents a decentralized blockchain network 600 where an NFT token is distributed and validated across multiple nodes in a peer-to-peer (P2P) system. The arrows indicate bidirectional communication and transaction validation between nodes. FIG. 6 consists of multiple nodes labelled as Node 1, Node 2, Node 3, ..., Node N. These nodes represent independent participants in the blockchain network that store, validate, and communicate transaction data. The nodes may be computers, servers, or any computing devices participating in the blockchain. The NFT token is transmitted between nodes. Each node verifies and validates the NFT transaction before adding it to the distributed ledger. The token remains immutable, ensuring security and ownership verification. For example, a digital collectible NFT (such as a unique artwork or a behavioural token) is issued to Node 1. Node 1 shares the transaction with other nodes for validation. If a consensus is reached, the NFT transaction is recorded in the blockchain ledger. Each node connects with multiple other nodes, forming a mesh-like communication structure. This ensures that no single entity has control, making the computing system 102 decentralized and resilient. The network likely follows a consensus mechanism, such as: Proof of Work (PoW), Proof of Stake (PoS) , Delegated Proof of Stake (DPoS). For example, if Node 1 initiates an NFT transaction, Nodes 2, 3, 4, N-1, and N verify the transaction. Once a consensus is reached, the NFT is recorded in the blockchain, and its ownership is transferred securely. Since each node maintains a copy of the blockchain ledger, the computing system 102 prevents fraud or tampering. Any attempt to modify the NFT token transaction would require modifying every node’s ledger, which is computationally infeasible. Blockchain encryption ensures data integrity, authenticity, and transparency.
[00209] FIG. 7 illustrates a process flow diagram representation of an exemplary method 700 for managing behaviour-based personalized services for end users 116, in accordance with an embodiment of the present disclosure. At step 702, the method 700 includes, periodically obtaining, by a processor 202, a multi-modal data associated with a user 116 from a plurality of data sources 106A-N. The multi-modal data comprises a behavioural data , an emotional data, a demographic data, and an ethnographic data. At step 704, the method 700 includes, determining, by the processor 202, user emotions and behavioural attributes associated with the user 116 based on the obtained multi-modal data using artificial-intelligence model. The behavioural attributes comprises a user interaction frequency, time spent on features, and content preferences. The user emotions comprises a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations, and wherein the behavioural attributes comprise at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies, and wherein the demographic and ethnographic factors comprise a location, a cultural background, an economic status, and language preferences of the user, and wherein the attitudinal interactions comprise analysing user decisions, preferences, and emotional responses during digital interactions.
[00210] Further, at step 706, the method 700 includes, generating, by the processor 202, a user persona model based on the determined user emotions and the behavioural attributes associated with the user. The user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user. At step 708, the method 700 includes, computing, by the processor 202, a behavioural score for the user 116 based on the generated user persona model. The behavioural score corresponds to a unique value assigned to the user. At step 710, the method 700 includes, predicting, by the processor 202, one or more probable interactions of the user 116 with one or more applications based on computed behavioural score using generative artificial intelligence (Gen-AI) models. At step 712, the method 700 includes, determining, by the processor 202, one or more modifications to be made to service configurations of the one or more applications by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications. The one or more virtual applications correspond to computer simulated version of the one or more applications. Furthermore, at step 714, the method 700 includes, modifying, by the processor 202, the service configurations of the one or more applications based on the determined one or more modifications. The service configurations comprises at least one of a network throughput, a response time, and inter-service links. Additionally, the method 700 includes outputting, by the processor 202, the computed behavioural score as a token to the one or more applications upon modifying the service configurations of the one or more applications.
[00211] The method 700 further includes, encrypting, by the processor 202, the computed behavioural score using one or more cryptographic models. Further, the method 700 includes, integrating, by the processor 202, the encrypted behavioural score as the token into the one or more applications using a blockchain network upon modifying the service configurations of the one or more applications. The token corresponds to a non-fungible token (NFT).
[00212] Further, the method 700 includes, validating, by the processor 202, the predicted one or more probable interactions and the computed behavioural score using historical interaction logs and artificial intelligence model and generating, by the processor 202, one or more recommendations for modifying the service configurations based on results of validation.
[00213] In determining the user emotions and the behavioural attributes associated with the user 116 based on the obtained multi-modal data using the artificial-intelligence model, the method 700 includes receiving, by the processor 202, one or more user responses to a series of visual and interactive interfaces. The visual and interactive interfaces comprises generative artificial intelligence models. Further, the method 700 includes identifying, by the processor 202, macro-level behavioural patterns of the user 116 based on the received one or more user responses. Furthermore, the method 700 includes generating, by the processor 202, behavioural patterns of the user 116 based on the identified macro-level behavioural patterns. Further, the method 700 includes determining, by the processor 202, the user emotions and the behavioural attributes associated with the user 116 based on the generated behavioural patterns of the user.
[00214] In computing the behavioural score for the user 116 based on the generated user persona model, the method 700 includes, classifying, by the processor 202, implicit and explicit behaviours of the users 116 based on external user data comprising visual inputs, heatmaps, chatbot interactions, and user engagement tracking. Further, the method 700 includes extracting, by the processor 202, demographic and ethnographic factors associated with the user 116 by performing attitudinal interactions with the user. Furthermore, the method 700 includes analysing, by the processor 202, the behavioural attributes and the user emotions exhibited by the user during the attitudinal interactions. Furthermore, the method 700 includes computing, by the processor 202, the behavioural score for the user 116 based on the analysed behavioural attributes and the user emotions.
[00215] Further, the method 700 includes continuously training, by the processor 202, the user persona model based on user reactions to recommendations, real-time interactions of the user 116 with the one or more applications, and user activity across the one or more applications;. Further, the method 700 includes continuously updating, by the processor 202, user profiles and recommendations through a feedback loop based on the trained user persona model. Furthermore, the method 700 includes dynamically adjusting, by the processor 202, the computed behavioural score based on the updated user profiles and the recommendations.
[00216] In modifying the service configurations of the one or more applications based on the determined one or more modifications, the method 700 includes, identifying, by the processor 202, at least one application among the one or more applications intended to be accessed by the user 116 based on the determined user emotions and the behavioural attributes. Further, the method 700 includes determining, by the processor 202, at least one service configuration of the identified at least one application relevant to the user 116 based on the computed behavioural score and the predicted future interactions. Furthermore, the method 700 includes determining, by the processor 202, one or more modifications to be made to the determined at least one service configuration. Further, the method 700 includes modifying, by the processor 202, the at least one service configuration based on the determined one or more modifications.
[00217] In predicting the one or more probable interactions of the user 116 with the one or more applications based on the computed behavioural score using the generative artificial intelligence (Gen-AI) models, the method 700 includes detecting, by the processor 202, sequential interaction patterns of the user 116 by analysing the behavioural score using a transformer-based neural network model. Further, the method 700 includes generating, by the processor 202, a set of probable user interactions with one or more applications based on the detected sequential interaction patterns, a prior engagement history, and inferred preferences. Furthermore, the method 700 includes ranking, by the processor 202, the set of probable user interactions using a probabilistic scoring model based on confidence levels derived from the detected sequential interaction patterns and real-time engagement metrics. Additionally, the method 700 includes tuning, by the processor 202, the ranked probable user interactions by incorporating multi-modal inputs comprising at least one of a text data, a voice data, a visual data, a biometric feedback, and an environmental context. Moreover, the method 700 includes simulating, by the processor 202, the tuned probable user interactions in a virtual environment using synthetic user profiles to validate model accuracy and response variations. Also, the method 700 includes predicting, by the processor 202, the one or more probable interactions of the user 116 with the one or more applications based on results of simulation.
[00218] In generating the user persona model based on the determined user emotions and the behavioural attributes associated with the user, the method 700 includes determining, by the processor 202, the user emotions based on a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations. Further, the method 700 includes identifying, by the processor 202, the behavioural attributes comprising at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies. Furthermore, the method 700 includes establishing, by the processor 202, a correlation between the determined user emotions and the behavioural attributes using a machine learning model trained on a real-time user data and a historical user data. Also, the method 700 includes periodically updating, by the processor 202, the user persona model based on real-time user interactions to indicate changes in the user emotions and the behavioural attributes. Furthermore, the method 700 includes configuring, by the processor 202, a structure of the user persona model using a nonlinear profiling model to represent the user emotions and the behavioural attributes. Also, the method 700 includes generating, by the processor 202, a graphical representation of the structured user personal model using a weighted graph structure. The graphical representation comprises one or more nodes corresponding to emotional states and the behavioural attributes. The graphical representation comprises one or more edges corresponding to a correlation strength value between the one or more nodes.
[00219] Furthermore, the method 700 includes generating, by the processor 202, one or more large language model (LLM) prompts based on the behavioural score and predicted user interactions to tune a user engagement level through a conversational AI. Further, the method 700 includes simulating, by the processor 202, the predicted user interactions within a virtual environment using the Gen-AI model and the generated one or more LLM prompts to assess response patterns and tune interaction predictions prior to deployment. Furthermore, the method 700 includes generating, by the processor 202, one or more LLM responses for the generated one or more LLM prompts using the Gen-AI model based on outcomes of the simulated user interactions. The one or more LLM responses comprise one or more recommendations on user interactions, service configurations, user behavioural patterns, the behavioural attributes and the user emotions.
[00220] Furthermore, the method 700 includes generating, by the processor 202, a user token based on the computed behavioural score. The behavioural score is computed from the user persona model representing correlations between the user emotions and the behavioural attributes. Further, the method 700 includes assigning, by the processor 202, a cryptographic key pair to the generated user token using a cryptographic model. Furthermore, the method 700 includes encrypting, by the processor 202, the user token using a blockchain-based encryption model upon assigning the cryptographic key pair. Further, the method 700 includes transmitting, by the processor 202, the encrypted user token to a distributed ledger network comprising a plurality of transactional nodes. Also, the method 700 includes validating, by the processor 202, the encrypted user token using a consensus technique between the plurality of transactional nodes and recording, by the processor 202, the encrypted user token onto the distributed ledger network as the non-fungible token (NFT). Additionally, the method 700 includes verifying, by the processor 202, one or more token access requests received from the plurality of transactional nodes using a cryptographic authentication model. The one or more token access requests comprises a cryptographic key for retrieving and decrypting the user token. Further, the method 700 includes controlling, by the processor 202, access to the encrypted user token based on results of verification using smart contract-based permissions. The smart contract-based permissions comprise one or more conditions for token retrieval, decryption, and transfer.
[00221] The present invention provides several technical advantages in securely managing, analysing, and utilizing user behavioural data through blockchain-based encryption, AI-driven behavioural modelling, and smart contract-controlled access. By integrating advanced cryptographic models, machine learning techniques, and distributed ledger technology, the computing system 102 enhances data security, privacy, and predictive accuracy while enabling seamless interoperability across multiple applications.
[00222] One of the key technical advantages of this invention is secure storage and transmission of user behavioural data using blockchain encryption. The computed behavioural score and the user persona model are encapsulated within a non-fungible token (NFT), which is encrypted using asymmetric cryptographic techniques. This prevents unauthorized access, tampering, or data breaches, ensuring that only verified entities with the appropriate cryptographic key may retrieve and decrypt the token. Immutable and tamper-proof storage on a distributed ledger network ensures data integrity. Decentralized token management eliminates reliance on a single entity, reducing security risks. Public-key cryptography and Zero-Knowledge Proofs (ZKP) enable controlled data sharing without exposing sensitive information.
[00223] The invention leverages AI and machine learning models to analyse user emotions, preferences, and behavioural patterns dynamically. By processing multi-modal inputs (text, speech, visual data, biometric signals, and the like), the computing system 102 creates a user persona model that continuously updates based on real-time interactions. Transformer-based neural networks (such as BERT or GPT) improve the accuracy of predicting future user interactions. Reinforcement learning fine-tunes the behavioural scoring model, improving personalization. Real-time adaptation to changes in user behaviour ensures dynamic service configuration modifications.
[00224] A major advantage is the use of smart contracts to enforce data access conditions. The computing system 102 implements role-based access control (RBAC) and compliance-driven retrieval policies to regulate who may access, decrypt, and transfer the behavioural token. Automated permission verification reduces reliance on centralized authentication servers. Regulatory compliance enforcement (e.g., GDPR, CCPA) ensures lawful processing of behavioural data. Time-restricted and purpose-specific decryption prevents long-term tracking or misuse.
[00225] The invention utilizes Generative AI models and transformer-based architectures to predict probable user interactions across applications. By analysing historical behaviour logs and engagement metrics, , the computing system 102 generates a ranked list of future interactions, optimizing service configurations dynamically. Predictive analytics improve service personalization, reducing latency and improving user experience. Self-learning AI models dynamically adapt based on new interactions, ensuring continuous improvement. Multi-modal data fusion (text, voice, and biometric) enhances prediction accuracy.
[00226] The invention provides a tokenized approach to store and manage behavioural scores on a blockchain network. By converting behavioural data into NFTs (Non-Fungible Tokens), the computing system 102 ensures secure interoperability between different services while preserving data privacy and ownership. Decentralized token storage prevents unauthorized modifications or deletions. Cross-application data portability allows users 116 to transfer their behavioural scores securely. Encrypted token-based transactions eliminate exposure of raw user data to third parties.
[00227] The computing system 102 automatically modifies service configurations (network throughput, response times, inter-service links, and the like) based on computed behavioural scores and predicted interactions. A virtual simulation environment validates changes before real-world deployment, ensuring optimized performance. Self-optimizing service orchestration reduces computational overhead and latency. AI-based configuration tuning enhances efficiency and resource allocation. Simulation-driven validation ensures performance improvements without service disruptions.
[00228] The computing system 102 leverages AI and machine learning models to analyse user emotions, preferences, and behavioural patterns dynamically. To reduce algorithmic biases, it incorporates bias correction techniques, fairness-aware training datasets, and explainable AI (XAI) models to ensure equitable outcomes. The AI models use adversarial debiasing techniques to detect and minimize unintentional biases in behavioural scoring. The computing system 102 applies re-weighting methods to ensure balanced representation across different demographic groups. The invention provides interpretable decision-making to identify and correct biased AI predictions in user interaction analysis. The AI models are continuously updated with bias monitoring frameworks, adjusting scoring mechanisms to prevent unfair penalization of certain user behaviours.
[00229] The written description describes the subject matter herein to enable any person skilled in the art to make and use the embodiments. The scope of the subject matter embodiments is defined by the claims and may include other modifications that occur to those skilled in the art. Such other modifications are intended to be within the scope of the claims if they have similar elements that do not differ from the literal language of the claims or if they include equivalent elements with insubstantial differences from the literal language of the claims.
[00230] The embodiments herein may include hardware and software elements. The embodiments that are implemented in software include but are not limited to, firmware, resident software, microcode, and the like. The functions performed by various modules described herein may be implemented in other modules or combinations of other modules. For the purposes of this description, a computer-usable or computer readable medium may be any apparatus that may include, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
[00231] The medium may be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium. Examples of a computer-readable medium include a semiconductor or solid-state memory, magnetic tape, a removable computer diskette, a random-access memory (RAM), a read-only memory (ROM), a rigid magnetic disk and an optical disk. Current examples of optical disks include compact disk-read only memory (CD-ROM), compact disk-read/write (CD-R/W) and DVD.
[00232] Input/output (I/O) devices (including but not limited to keyboards, displays, pointing devices, and the like.) may be coupled to the computing system 102 either directly or through intervening I/O controllers. Network adapters may also be coupled to the computing system 102 to enable the data processing system to become coupled to other data processing systems or remote printers or storage devices through intervening private or public networks. Modems, cable modem and Ethernet cards are just a few of the currently available types of network adapters.
[00233] A representative hardware environment for practicing the embodiments may include a hardware configuration of an information handling/computer system in accordance with the embodiments herein. The computing system 102 herein includes at least one processor or central processing unit (CPU). The CPUs are interconnected via system bus 208 to various devices such as a random-access memory (RAM), read-only memory (ROM), and an input/output (I/O) adapter. The I/O adapter may connect to peripheral devices, such as disk units and tape drives, or other program storage devices that are readable by the computing system 102. The computing system 102 may read the inventive instructions on the program storage devices and follow these instructions to execute the methodology of the embodiments herein.
[00234] The computing system 102 further includes a user interface adapter that connects a keyboard, mouse, speaker, microphone, and/or other user interface-devices such as a touch screen device (not shown) to the bus to gather user input. Additionally, a communication adapter connects the bus to a data processing network, and a display adapter connects the bus to a display device which may be embodied as an output device such as a monitor, printer, or transmitter, for example.
[00235] A description of an embodiment with several components in communication with each other does not imply that all such components are required. On the contrary, a variety of optional components are described to illustrate the wide variety of possible embodiments of the invention. When a single device or article is described herein, it will be apparent that more than one device/article (whether or not they cooperate) may be used in place of a single device/article. Similarly, where more than one device or article is described herein (whether or not they cooperate), it will be apparent that a single device/article may be used in place of the more than one device or article, or a different number of devices/articles may be used instead of the shown number of devices or programs. The functionality and/or the features of a device may be alternatively embodied by one or more other devices which are not explicitly described as having such functionality/features. Thus, other embodiments of the invention need not include the device itself.
[00236] The illustrated steps are set out to explain the exemplary embodiments shown, and it should be anticipated that ongoing technological development will change the manner in which particular functions are performed. These examples are presented herein for purposes of illustration, and not limitation. Further, the boundaries of the functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternative boundaries may be defined so long as the specified functions and relationships thereof are appropriately performed. Alternatives (including equivalents, extensions, variations, deviations, and the like., of those described herein) will be apparent to persons skilled in the relevant art(s) based on the teachings contained herein. Such alternatives fall within the scope and spirit of the disclosed embodiments. Also, the words "comprising," "having," "containing," and "including," and other similar forms are intended to be equivalent in meaning and be open-ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items or meant to be limited to only the listed item or items. It must also be noted that as used herein and in the appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise.
[00237] Finally, the language used in the specification has been principally selected for readability and instructional purposes, and it may not have been selected to delineate or circumscribe the inventive subject matter. It is therefore intended that the scope of the invention be limited not by this detailed description, but rather by any claims that issue on an application based here on. Accordingly, the embodiments of the present invention are intended to be illustrative, but not limiting, of the scope of the invention, which is set forth in the following claims.
,CLAIMS:WE CLAIM:
1. A system (102) for managing behaviour-based personalized services for end users (116) in a computing environment (100a), the system (102) comprising:
a processor (202);
a memory (204) coupled to the processor (202), wherein the memory (204) comprises processor-executable instructions, which on execution, cause the processor (202) to:
periodically obtain a multi-modal data associated with a user (116) from a plurality of data sources (106A-N), wherein the multi-modal data comprises a behavioural data, an emotional data, a demographic data, and an ethnographic data;
determine user emotions and behavioural attributes associated with the user (116) based on the obtained multi-modal data using artificial-intelligence model, wherein the behavioural attributes comprises a user interaction frequency, time spent on features, and content preferences;
generate a user persona model based on the determined user emotions and the behavioural attributes associated with the user (116), wherein the user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user (116);
compute a behavioural score for the user (116) based on the generated user persona model, wherein the behavioural score corresponds to a unique value assigned to the user;
predict one or more probable interactions of the user (116) with one or more applications based on computed behavioural score using generative artificial intelligence (Gen-AI) models;
determine one or more modifications to be made to service configurations of the one or more applications (108A-N) by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications, wherein the one or more virtual applications correspond to computer simulated version of the one or more applications (108A-N);
modify the service configurations of the one or more applications (108A-N) based on the determined one or more modifications, wherein the service configurations comprises at least one of a network throughput, a response time, and inter-service links; and
output the computed behavioural score as a token to the one or more applications (108A-N) upon modifying the service configurations of the one or more applications (108A-N).

2. The system (102) as claimed in claim 1, wherein the processor (202) is further configured to:
encrypt the computed behavioural score using one or more cryptographic models; and
integrate the encrypted behavioural score as the token into the one or more applications (108A-N) using a blockchain network upon modifying the service configurations of the one or more applications (108A-N), wherein the token corresponds to a non-fungible token (NFT).

3. The system (102) as claimed in claim 1, wherein the processor (202) is configured to:
validate the predicted one or more probable interactions and the computed behavioural score using historical interaction logs and artificial intelligence model; and
generate one or more recommendations for modifying the service configurations based on results of validation.

4. The system (102) as claimed in claim 1, wherein the user emotions comprises a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations, and wherein the behavioural attributes comprise at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies, and wherein the demographic and ethnographic factors comprise a location, a cultural background, an economic status, and language preferences of the user, and wherein the attitudinal interactions comprise analysing user decisions, preferences, and emotional responses during digital interactions.

5. The system (102) as claimed in claim 1, wherein to determine the user emotions and the behavioural attributes associated with the user (116) based on the obtained multi-modal data using the artificial-intelligence model, the processor (202) is configured to:
receive one or more user responses to a series of visual and interactive interfaces, wherein the visual and interactive interfaces comprises generative artificial intelligence models;
identify macro-level behavioural patterns of the user (116) based on the received one or more user responses;
generate behavioural patterns of the user (116) based on the identified macro-level behavioural patterns; and
determine the user emotions and the behavioural attributes associated with the user (116) based on the generated behavioural patterns of the user (116).

6. The system (102) as claimed in claim 1, wherein to compute the behavioural score for the user (116) based on the generated user persona model, the processor (202) is configured to:
classify implicit and explicit behaviours of the users (116) based on external user data comprising visual inputs, heatmaps, chatbot interactions, and user engagement tracking;
extract demographic and ethnographic factors associated with the user (116) by performing attitudinal interactions with the user (116);
analyse the behavioural attributes and the user emotions exhibited by the user (116) during the attitudinal interactions; and
compute the behavioural score for the user (116) based on the analysed behavioural attributes and the user emotions.

7. The system (102) as claimed in claim 6, wherein the processor (202) is configured to:
continuously train the user persona model based on user reactions to recommendations, real-time interactions of the user (116) with the one or more applications (108A-N), and user activity across the one or more applications (108A-N);
continuously update user profiles and recommendations through a feedback loop based on the trained user persona model; and
dynamically adjust the computed behavioural score based on the updated user profiles and the recommendations.

8. The system (202) as claimed in claim 1, wherein to modify the service configurations of the one or more applications (108A-N) based on the determined one or more modifications, the processor (202) is configured to:
identify at least one application among the one or more applications (108A-N) intended to be accessed by the user (116) based on the determined user emotions and the behavioural attributes;
determine at least one service configuration of the identified at least one application relevant to the user (116) based on the computed behavioural score and the predicted future interactions;
determine one or more modifications to be made to the determined at least one service configuration; and
modify the at least one service configuration based on the determined one or more modifications.

9. The system (102) as claimed in claim 1, wherein to predict the one or more probable interactions of the user (116) with the one or more applications (108A-N) based on the computed behavioural score using the generative artificial intelligence (Gen-AI) models, the processor (202) is configured to:
detect sequential interaction patterns of the user (116) by analysing the behavioural score using a transformer-based neural network model;
generate a set of probable user interactions with one or more applications (108A-N) based on the detected sequential interaction patterns, a prior engagement history, and inferred preferences;
rank the set of probable user interactions using a probabilistic scoring model based on confidence levels derived from the detected sequential interaction patterns and real-time engagement metrics;
tune the ranked probable user interactions by incorporating multi-modal inputs comprising at least one of a text data, a voice data, a visual data, a biometric feedback, and an environmental context;
simulate the tuned probable user interactions in a virtual environment using synthetic user profiles to validate model accuracy and response variations; and
predict the one or more probable interactions of the user (116) with the one or more applications (108A-N) based on results of simulation.

10. The system (102) as claimed in claim 1, wherein to generate the user persona model based on the determined user emotions and the behavioural attributes associated with the user (116), the processor (202) is configured to:
determine the user emotions based on a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations;
identify the behavioural attributes comprising at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies;
establish a correlation between the determined user emotions and the behavioural attributes using a machine learning model trained on a real-time user data and a historical user data;
periodically update the user persona model based on real-time user interactions to indicate changes in the user emotions and the behavioural attributes;
structure the user persona model using a nonlinear profiling model to represent the user emotions and the behavioural attributes; and
generate a graphical representation of the structured user personal model using a weighted graph structure, wherein the graphical representation comprises one or more nodes corresponding to emotional states and the behavioural attributes, and wherein the graphical representation comprises one or more edges corresponding to a correlation strength value between the one or more nodes.

11. The system (102) as claimed in claim 1, wherein the processor (202) is further configured to:
generate one or more large language model (LLM) prompts based on the behavioural score and predicted user interactions to tune a user engagement level through a conversational AI;
simulate the predicted user interactions within a virtual environment using the Gen-AI model and the generated one or more LLM prompts to assess response patterns and tune interaction predictions prior to deployment; and
generate one or more LLM responses for the generated one or more LLM prompts using the Gen-AI model based on outcomes of the simulated user interactions, wherein the one or more LLM responses comprise one or more recommendations on user interactions, service configurations, user behavioural patterns, the behavioural attributes and the user emotions.

12. The system (102) as claimed in claim 1, wherein the processor (202) is further configured to:
generate a user token based on the computed behavioural score, wherein the behavioural score is computed from the user persona model representing correlations between the user emotions and the behavioural attributes;
assign a cryptographic key pair to the generated user token using a cryptographic model;
encrypt the user token using a blockchain-based encryption model upon assigning the cryptographic key pair;
transmit the encrypted user token to a distributed ledger network comprising a plurality of transactional nodes;
validate the encrypted user token using a consensus technique between the plurality of transactional nodes;
record the encrypted user token onto the distributed ledger network as the non-fungible token (NFT);
verify one or more token access requests received from the plurality of transactional nodes using a cryptographic authentication model, wherein the one or more token access requests comprises a cryptographic key for retrieving and decrypting the user token; and
control access to the encrypted user token based on results of verification using smart contract-based permissions, wherein the smart contract-based permissions comprise one or more conditions for token retrieval, decryption, and transfer.

13. A method (700) for managing behaviour-based personalized services for end users (116) in a computing environment (100a), the method (700) comprising:
periodically obtaining, by a processor (202), a multi-modal data associated with a user (116) from a plurality of data sources (108A-N), wherein the multi-modal data comprises a behavioural data , an emotional data, a demographic data, and an ethnographic data;
determining, by the processor (202), user emotions and behavioural attributes associated with the user (116) based on the obtained multi-modal data using artificial-intelligence model, wherein the behavioural attributes comprises a user interaction frequency, time spent on features, and content preferences;
generating, by the processor (202), a user persona model based on the determined user emotions and the behavioural attributes associated with the user (116), wherein the user persona model represents a correlation between the determined user emotions and the behavioural attributes of the user (116);
computing, by the processor (202), a behavioural score for the user (116) based on the generated user persona model, wherein the behavioural score corresponds to a unique value assigned to the user;
predicting, by the processor (202), one or more probable interactions of the user (116) with one or more applications based on computed behavioural score using generative artificial intelligence (Gen-AI) models;
determining, by the processor (202), one or more modifications to be made to service configurations of the one or more applications (108A-N) by simulating the computed behavioural score and the predicted future interactions onto one or more virtual applications, wherein the one or more virtual applications correspond to computer simulated version of the one or more applications (108A-N);
modifying, by the processor (202), the service configurations of the one or more applications (108A-N) based on the determined one or more modifications, wherein the service configurations comprises at least one of a network throughput, a response time, and inter-service links; and
outputting, by the processor (202), the computed behavioural score as a token to the one or more applications (108A-N) upon modifying the service configurations of the one or more applications (108A-N).

14. The method (700) as claimed in claim 13, further comprising:
encrypting, by the processor (202), the computed behavioural score using one or more cryptographic models; and
integrating, by the processor (202), the encrypted behavioural score as the token into the one or more applications (108A-N) using a blockchain network upon modifying the service configurations of the one or more applications (108A-N), wherein the token corresponds to a non-fungible token (NFT).

15. The method (700) as claimed in claim 13, further comprising:
validating, by the processor (202), the predicted one or more probable interactions and the computed behavioural score using historical interaction logs and artificial intelligence model; and
generating, by the processor (202), one or more recommendations for modifying the service configurations based on results of validation.

16. The method (700) as claimed in claim 13, wherein the user emotions comprises a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations, and wherein the behavioural attributes comprise at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies, and wherein the demographic and ethnographic factors comprise a location, a cultural background, an economic status, and language preferences of the user, and wherein the attitudinal interactions comprise analysing user decisions, preferences, and emotional responses during digital interactions.

17. The method (700) as claimed in claim 13, wherein determining the user emotions and the behavioural attributes associated with the user (116) based on the obtained multi-modal data using the artificial-intelligence model comprises:
receiving, by the processor (202), one or more user responses to a series of visual and interactive interfaces, wherein the visual and interactive interfaces comprises generative artificial intelligence models;
identifying, by the processor (202), macro-level behavioural patterns of the user (116) based on the received one or more user responses;
generating, by the processor (202), behavioural patterns of the user (116) based on the identified macro-level behavioural patterns; and
determining, by the processor (202), the user emotions and the behavioural attributes associated with the user (116) based on the generated behavioural patterns of the user (116).

18. The method (700) as claimed in claim 13, wherein computing the behavioural score for the user (116) based on the generated user persona model comprises:
classifying, by the processor (202), implicit and explicit behaviours of the users (116) based on external user data comprising visual inputs, heatmaps, chatbot interactions, and user engagement tracking;
extracting, by the processor (202), demographic and ethnographic factors associated with the user (116) by performing attitudinal interactions with the user (116);
analysing, by the processor (202), the behavioural attributes and the user emotions exhibited by the user (116) during the attitudinal interactions; and
computing, by the processor (202), the behavioural score for the user (116) based on the analysed behavioural attributes and the user emotions.

19. The method (700) as claimed in claim 18, further comprising:
continuously training, by the processor (202), the user persona model based on user reactions to recommendations, real-time interactions of the user (116) with the one or more applications (108A-N), and user activity across the one or more applications (108A-N);
continuously updating, by the processor (202), user profiles and recommendations through a feedback loop based on the trained user persona model; and
dynamically adjusting, by the processor (202), the computed behavioural score based on the updated user profiles and the recommendations.

20. The method (700) as claimed in claim 13, wherein modifying the service configurations of the one or more applications (108A-N) based on the determined one or more modifications comprises:
identifying, by the processor (202), at least one application among the one or more applications (108A-N) intended to be accessed by the user (116) based on the determined user emotions and the behavioural attributes;
determining, by the processor (202), at least one service configuration of the identified at least one application relevant to the user (116) based on the computed behavioural score and the predicted future interactions;
determining, by the processor (202), one or more modifications to be made to the determined at least one service configuration; and
modifying, by the processor (202), the at least one service configuration based on the determined one or more modifications.

21. The method (700) as claimed in claim 13, wherein predicting the one or more probable interactions of the user (116) with the one or more applications (108A-N) based on the computed behavioural score using the generative artificial intelligence (Gen-AI) models, comprises:
detecting, by the processor (202), sequential interaction patterns of the user (116) by analysing the behavioural score using a transformer-based neural network model;
generating, by the processor (202), a set of probable user interactions with one or more applications (108A-N) based on the detected sequential interaction patterns, a prior engagement history, and inferred preferences;
ranking, by the processor (202), the set of probable user interactions using a probabilistic scoring model based on confidence levels derived from the detected sequential interaction patterns and real-time engagement metrics;
tuning, by the processor (202), the ranked probable user interactions by incorporating multi-modal inputs comprising at least one of a text data, a voice data, a visual data, a biometric feedback, and an environmental context;
simulating, by the processor (202), the tuned probable user interactions in a virtual environment using synthetic user profiles to validate model accuracy and response variations; and
predicting, by the processor (202), the one or more probable interactions of the user (116) with the one or more applications (108A-N) based on results of simulation.

22. The method (700) as claimed in claim 13, wherein generating the user persona model based on the determined user emotions and the behavioural attributes associated with the user (116) comprises:
determining, by the processor (202), the user emotions based on a sentiment analysis of user interactions, facial expressions, speech patterns, biometric signals, emotional stability, and mood variations;
identifying, by the processor (202), the behavioural attributes comprising at least one of user interaction frequency, engagement levels, response times, preferences, and decision-making tendencies;
establishing, by the processor (202), a correlation between the determined user emotions and the behavioural attributes using a machine learning model trained on a real-time user data and a historical user data;
periodically updating, by the processor (202), the user persona model based on real-time user interactions to indicate changes in the user emotions and the behavioural attributes;
configuring, by the processor (202), a structure of the user persona model using a nonlinear profiling model to represent the user emotions and the behavioural attributes; and
generating, by the processor (202), a graphical representation of the structured user personal model using a weighted graph structure, wherein the graphical representation comprises one or more nodes corresponding to emotional states and the behavioural attributes, and wherein the graphical representation comprises one or more edges corresponding to a correlation strength value between the one or more nodes.

23. The method (700) as claimed in claim 13, further comprising:
generating, by the processor (202), one or more large language model (LLM) prompts based on the behavioural score and predicted user interactions to tune a user engagement level through a conversational AI;
simulating, by the processor (202), the predicted user interactions within a virtual environment using the Gen-AI model and the generated one or more LLM prompts to assess response patterns and tune interaction predictions prior to deployment; and
generating, by the processor (202), one or more LLM responses for the generated one or more LLM prompts using the Gen-AI model based on outcomes of the simulated user interactions, wherein the one or more LLM responses comprise one or more recommendations on user interactions, service configurations, user behavioural patterns, the behavioural attributes and the user emotions.

24. The method (700) as claimed in claim 13, further comprising:
generating, by the processor (202), a user token based on the computed behavioural score, wherein the behavioural score is computed from the user persona model representing correlations between the user emotions and the behavioural attributes;
assigning, by the processor (202), a cryptographic key pair to the generated user token using a cryptographic model;
encrypting, by the processor (202), the user token using a blockchain-based encryption model upon assigning the cryptographic key pair;
transmitting, by the processor (202), the encrypted user token to a distributed ledger network comprising a plurality of transactional nodes;
validating, by the processor (202), the encrypted user token using a consensus technique between the plurality of transactional nodes;
recording, by the processor (202), the encrypted user token onto the distributed ledger network as the non-fungible token (NFT);
verifying, by the processor (202), one or more token access requests received from the plurality of transactional nodes using a cryptographic authentication model, wherein the one or more token access requests comprises a cryptographic key for retrieving and decrypting the user token; and

controlling, by the processor (202), access to the encrypted user token based on results of verification using smart contract-based permissions, wherein the smart contract-based permissions comprise one or more conditions for token retrieval, decryption, and transfer.