Description:FIELD OF THE INVENTION
[001] The present invention generally relates to food science and technology, and more specifically to systems and methods for recommending culinary recipes based on user requirements, such as dietary preferences, nutritional needs, and health conditions.

BACKGROUND OF THE INVENTION
[002] In recent years, with the increasing awareness of health and nutrition, there has been an increasing demand for personalised nutrition. Consumers are becoming increasingly conscious regarding their food choices, and are seeking meals that align with their health goals, dietary restrictions, and lifestyle preferences. Whether for weight management, disease prevention, or athletic performance, individuals now require more than just generic nutritional advice. Instead, there is a need for personalized recommendations that take into account the unique needs of each individual.
[003] Traditional food/ culinary recommendation systems often rely on broad categories or simple processes and methods based on limited factors, such as calorie count or basic macronutrient balance. While these systems can offer some level of customization, they fail to address the full spectrum of user requirements. Many factors contribute to an individual's dietary needs, specific medical conditions, individual taste preferences and the like. As a result, existing systems often fall short in providing truly personalized food recipes that can comprehensively satisfy all these diverse and sometimes conflicting needs.
[004] There is a growing interest in using data-driven approaches to create more precise and effective food recommendations. By leveraging advances in artificial intelligence (AI), machine learning, and nutrition science, it is possible to develop systems capable of analysing large datasets, identifying patterns, and offering personalized food recipes.
[005] Existing methods for recommending food/ culinary recipes fails to effectively integrate diverse factors, leading to suboptimal food choices that may not fully align with the user’s health objectives or preferences. The conventional approaches to addressing these challenges include simple heuristics or rules of thumb that can be learned over time. Additionally, user-specific food preferences, including dietary restrictions and allergies, further complicate cooking process. A food preparer may need to modify recipes to accommodate these preferences but might not have an expertise to do so successfully. For instance, substituting a dairy product for a non-dairy alternative requires an understanding of how these substitutes will interact with other ingredients during cooking. Moreover, conducting such research for the substitution requires a certain level of preexisting knowledge about what to look for and the research is time consuming.
[006] Thus, there is a need in the art for systems and methods for recommending culinary recipes based on user requirements, which addresses at least the aforementioned problems.
SUMMARY OF THE INVENTION
[007] The present invention in one aspect is directed to a system for recommending culinary recipes based on user requirements and user information, the system has one or more processors and a memory. The memory is operatively coupled to one or more processors. The memory includes a set of instructions which, when executed by the processor, cause the processor to receive data for a primary culinary recipe of a culinary item corresponding to user requirements. The processor is also configured to determine distribution of a plurality of ingredients for the primary culinary recipe of the culinary item, whereby the distribution is dynamic across a plurality of culinary groups, inferring a distinct weightage associated with the plurality of culinary groups.
[008] As per the invention, the processor is further configured to calculate statistical measures for a nutritional value of each ingredient with a weighted mean, standard deviation, and confidence intervals, to estimate a probability of predefined range of the plurality of ingredients. Further, the processor is configured to train a Machine Learning (ML) model using the statistical measures for one or more nutritional values, whereby the trained ML model is configured to predict one or more recipes with one or more combinations of ingredients to achieve a required nutritional value. Furthermore, the processor is configured to map one or more nutritional values of each of the plurality of culinary ingredients form one or more predicted recipes with nutritional value of the primary culinary recipe using an artificial intelligence (AI) driven feature selection technique. Moreover, the processor is configured to recommend a preferred culinary recipe by selecting a combination of ingredients from the predicted one or more combinations of ingredients with an optimized proportion of the plurality of ingredients for obtaining the required nutritional value of the culinary item.
[009] In an embodiment, the processor is configured to analyze relationships between the predicted one or more combinations of ingredients for formulation the culinary item corresponding to the predefined range of the plurality of ingredients.
[010] In another embodiment, the processor is configured to detect anomalies in the relationships for mapping one or more nutritional values of each of the plurality of culinary ingredients.
[011] In a further embodiment of the present invention, the processor is configured to optimize proportions of one or more combinations of the of predicted plurality of ingredients through an AI-driven optimization technique to validate the nutritional consistency in the culinary item.
[012] In yet another embodiment of the present invention, the processor is configured to refine the optimization of the plurality of ingredients based on feedback from the predictions of the ML model, whereby the ML model utilizes active learning for obtaining accuracy of the ML model.
[013] In a further embodiment of the present invention, the processor is configured to clean and preprocess the data to handle missing values, normalize nutritional values, and transform a categorical data corresponding to the plurality of ingredients into usable numerical features.
[014] In yet another embodiment of the present invention, one or more AI-driven distance matrices configured to analyze similarity between the plurality of ingredients and ability to substitute with another ingredient while maintaining flavor, texture, and nutritional value of the culinary item.
[015] Furthermore, in an embodiment of the present invention, the AI-driven feature selection technique is configured to identify at least one of the features including taste, texture, and cost along with the nutritional value.
[016] Furthermore, in another embodiment of the present invention, the received data comprises a plurality of primary culinary recipes with a plurality of ingredients and a set of features responsible for the preferred culinary recipe.
[017] Furthermore, in an embodiment of the present invention, the plurality of instructions causes the processor to analyze relationship between the predicted one or more combinations of ingredients by constructing and utilizing the one or more AI-driven distance matrices to quantify clustering of the plurality of culinary groups.
[018] Moreover, in an embodiment of the present invention, the plurality of ingredients with nutritional data is extracted from a database by using a combination of Topic Modeling and Knowledge Graphs (KGs)technique, thereby ensuring the extracted data is structured, whereby the database is hosted on a server.
[019] Moreover, in an embodiment of the present invention, the analysis module is configured to correlate the predicted features with the nutritional requirements, sensory requirements, and functional requirements of the primary culinary recipe.
[020] Moreover, in an embodiment of the present invention, the weighted mean is based on average of the plurality of culinary groups sharing a same type of the culinary item.
[021] In another aspect, the present invention relates to a method for recommending culinary recipes based on user requirements. The method of the present invention includes receiving, data for a primary culinary recipe for a culinary item and corresponding user requirements. The method of the present invention also includes determining, distribution of the plurality of ingredients for the primary culinary recipe for the culinary item, whereby the distribution is dynamic across a plurality of culinary groups, inferring a distinct weightage associated with the plurality of culinary groups. Further, the method of the present invention includes calculating, statistical measures for a nutritional value of each ingredient with a weighted mean, standard deviation, and confidence intervals, to estimate a probability of predefined range of the plurality of ingredients. Furthermore, the method of the present invention includes training, a machine learning (ML) model using the statistical measures for one or more nutritional values, whereby the trained ML model is configured to predict one or more recipes with one or more combinations of ingredients to achieve a required nutritional value. Moreover, the method of the present invention includes mapping, one or more nutritional values of each of the plurality of culinary ingredients form one or more predicted recipes with nutritional value of the primary culinary recipe using an artificial intelligence (AI) driven feature selection technique. Moreover, the method of the present invention includes recommending, a preferred culinary recipe by selecting a combination of ingredients from the predicted one or more combinations of ingredients with an optimized proportion of the plurality of ingredients for obtaining the required nutritional value of the culinary item.
[022] In an embodiment of the present invention, the method includes applying, clustering techniques to group a plurality of ingredients with same nutritional and one or more sensory characteristics and identifying, outlier ingredients by using anomaly detection to identify an ingredient mismatched with the nutritional profile or sensory characteristics of the primary culinary recipe.
[023] In another embodiment of the present invention, the sensory characteristics comprises at least one of taste, smell, texture, appearance, and color.

BRIEF DESCRIPTION OF THE DRAWINGS
[024] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the description, serve to explain the disclosed principles. Reference has been made to embodiments of the invention, examples of which may be illustrated in accompanying figures. These figures constitute a part of this disclosure are intended to be illustrative, and together with the description, serve to explain the invention. Although the invention is generally described in context of these embodiments, it should be understood that it is not intended to limit the scope of the invention to these particular embodiments.
Figure 1 shows components of a system for recommending culinary recipes in accordance with an embodiment of the invention.
Figure 2 shows a network environment of the system for recommending culinary recipes in accordance with an embodiment of the invention.
Figure 3 shows steps involved in a method for recommending culinary recipes in accordance with an embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION
[025] Embodiments of the present disclosure, disclose a system and a method for recommending culinary recipes. The terms “comprises.... a”, “comprising”, or any other variations thereof used in the specification, are intended to cover a non-exclusive inclusion, such that a device that comprises a list of components or steps does not include only those components or steps but may include other components or steps not expressly listed or inherent to such device. In other words, one or more elements in an assembly proceeded by “comprises... a” does not, without more constraints, preclude the existence of other elements or additional elements in the device.
[026] For the purposes of promoting an understanding of the principles of the disclosure, reference will now be made to specific embodiments illustrated in the drawings and specific language will be used to describe the same. 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 methods, and such further applications of the principles of the disclosure as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention pertains.
[027] Referring to Figure 1, according to the invention, the system (100) includes one or more processors (102) and a memory (104). The one or more processors (102), in the context of this invention, may refer to a microprocessor or central processing unit (CPU) which is designed to execute instructions and perform computations within a computing system. The processor (102) is a primary component responsible for interpreting managing data flow, and handling input/output operations. The processor typically includes multiple cores that enable parallel processing, improving efficiency and performance. The one or more processors (102) operates by fetching, decoding, and executing instructions stored in the memory (104), performing arithmetic, logical, control, and input/output operations as needed. Architecture of the one or more processors (102) may incorporate various technologies such as multi-core configurations, specialized instruction sets, and integrated components to optimize processing power, speed, and energy consumption for specific applications or use cases.
[028] The memory (104) is operatively coupled to one or more processors (102). In an embodiment, the memory (104), as described in this invention, refers to a system component responsible for storing data (208) and program instructions that are accessed by the one or more processors (102) during operation. In another embodiment, the memory (104) includes both volatile memory, such as Random Access Memory (RAM), which temporarily stores data that is actively being used or processed, and non-volatile memory, such as Read-Only Memory (ROM) or flash storage. The present invention may incorporate novel configurations, methods, or architectures that enhance memory performance, data integrity, and energy efficiency to meet the demands of specific applications or computing environments. Further, the memory (104) includes a set of instructions which, when executed by the one or more processors (102), cause the one or more processors (102) to receive data (208) for a primary culinary recipe (110) of a culinary item corresponding to user requirements (112). In an embodiment, the primary culinary recipe (110) may be an original recipe of the culinary item based on which the system (100) may provide a preferred culinary recipe with same characteristics and high nutritional value.
[029] In an embodiment, the system (100) is configured to clean and preprocess the data to handle missing values, normalize nutritional values, and transform a categorical data corresponding to the plurality of ingredients (114) into usable numerical features. In an embodiment, the received data includes a plurality of primary culinary recipes with a plurality of ingredients (114) and a set of features responsible for the preferred culinary recipe. In an embodiment, the plurality of ingredients (114) with nutritional data is extracted from a database (116) by using a combination of topic modeling and knowledge graphs (KGs) technique. The topic modeling and knowledge graph technique is an advanced technique that enhances text understanding, knowledge discovery, and information retrieval. This hybrid approach integrates unsupervised ingredient extraction with structured semantic knowledge. technique thereby ensuring the extracted data is structured. The database (116) is hosted on a server (118). In an embodiment, the server (118) may a cloud server, database server, file server, web server, and the like. In an embodiment, the topic modelling is an unsupervised learning method that extracts latent topics from the plurality of groups of culinary recipes. In an embodiment, the knowledge graph is a structured representation of facts, concepts, and their relationships. Integrating the topic modelling and knowledge graphs offers several advantages such as enhanced interpretability of culinary recipes, improved coherence, context aware discovery of the plurality of ingredients, semantic enrichment, better downstream natural language processing (NLP) tasks. The one or more processors (102) are also configured to determine distribution of a plurality of ingredients (114) for the primary culinary recipe (110) of the culinary item. The distribution is dynamic across a plurality of culinary groups, inferring a distinct weightage associated with the plurality of culinary groups. In an embodiment, the allocation of ingredients is not fixed or static; the allocation is "dynamic," that is the distribution may change across various culinary groups, which may refer to different types of cuisines, cooking styles, or dietary preferences such as vegan, Italian, or gluten-free groups, and the like.
[030] In another embodiment, the distribution is also based on inferring or understanding the weightage (such as relative importance) of each culinary group. Each group might require different proportions of certain ingredients based on their unique characteristics or preferences. For example, a vegan group might have a different ratio of plant-based ingredients than a traditional animal-based group.
[031] Further, the one or more processors (102) are configured to calculate statistical measures for a nutritional value of each ingredient with a weighted mean, standard deviation, and confidence intervals, to estimate a probability of predefined range of the plurality of ingredients (114). The system (100) is configured to analyse the nutritional values of each of the ingredients in the primary culinary recipe (110). For example, the nutritional value may include calories, protein, fat content, vitamins, and the like. In an embodiment, the system (100) includes one or more AI-driven distance matrices configured to analyze similarity between the plurality of ingredients (114) and ability to substitute with another ingredient while maintaining flavor, texture, and nutritional value of the culinary item.
[032] In an embodiment, the AI-driven distance matrices are a type of distance or similarity matrix that uses AI techniques to measure the relationships or similarities between data points, in a context-aware manner. Unlike traditional distance metrics, which are often based on simple mathematical calculations, the AI-driven distance matrices use machine learning models (like neural networks or unsupervised learning techniques) to compute relationships between data points.
[033] The statistical measures may be applied to the nutritional values of the plurality of ingredients (114). In an embodiment, the weighted mean is a type of average where some data points (such as ingredients) are given more importance. For example, an ingredient that contributes more to the total weight of a dish may be assigned a higher weight in the mean calculation. In an embodiment, the standard deviation is a measure of the spread-out values. A higher standard deviation means high variation of the nutritional values of the ingredients, while a lower standard deviation means the values are more consistent.
[034] In an embodiment, the confidence intervals is a statistical range that may contains a true value. For example, while calculating the average calorie content of a recipe, the confidence interval may provide a range (such as 200–220 calories) within which the true average is likely to fall. In an embodiment, calculation of the confidence intervals may provide the precision for estimating a preferred range of the plurality of ingredients (114). In an embodiment, the predefined range may be a specific target or set of limits which may be added in a recipe. For example, the user requirement is to have total calories in between 300 and 400, or that the protein content falls between 10 and 20 grams in the culinary item. Here, 300 to 400 is a predefined range of total calories and 10 and 20 grams is a predefined range of the protein content.
[035] Furthermore, the one or more processors (102) are configured to train a machine learning (ML) model using the statistical measures for one or more nutritional values, whereby the trained ML model is configured to predict one or more recipes with one or more combinations of ingredients to achieve a required nutritional value. In an embodiment, the training of ML model involves gathering and preparing data, choosing a model, training the model to learn from the data, evaluating its performance, and improving the performance over time to make accurate predictions.
[036] In an embodiment, the ML model is trained using statistical measures such as averages, variances, or confidence intervals of nutritional values for various ingredients. In an embodiment, the ML model is configured to predict recipes that use combinations of ingredients which are required to meet a required nutritional value. In another embodiment, the ML model learns the relationships between different ingredients and their nutritional values. The trained ML model, may suggest right ingredients and proportions that, when combined, helps in achieving a desired nutritional outcome for the preferred culinary recipe.
[037] Furthermore, the one or more processors (102) are configured to map one or more nutritional values of each of the plurality of culinary ingredients form one or more predicted recipes with nutritional value of the primary culinary recipe (110) using an artificial intelligence (AI) driven feature selection technique. In an embodiment, the system (100) is configured to detect anomalies in the relationships for mapping one or more nutritional values of each of the plurality of culinary ingredients. In an embodiment, AI refers to the simulation of human intelligence processes by machines, particularly computers. AI involves creating systems or algorithms that can perform tasks typically requiring human intelligence, such as learning, reasoning, problem-solving, understanding language, and perception.
[038] In an embodiment, the AI-driven feature selection is a technique used to predict and select the most relevant features (variables or attributes) from a dataset to improve the performance of the ML model. In an embodiment, the one or more processors (102) are configured to correlate the predicted features with the nutritional requirements, sensory requirements, and functional requirements of the primary culinary recipe. In essence, the AI-driven feature selection technique helps to identify the most important input variables that uses the ML model's predictive feature while removing irrelevant or redundant features. AI-driven feature selection techniques use AI processes to automatically select the best subset of features, instead of relying solely on human judgment or simple statistical methods. These techniques can be applied to various machine learning models such as but not limited to reinforcement learning and neural networks, to enhance the accuracy of the output, reduce complexity, and speed up training time.
[039] In another embodiment, the system (100) is configured to optimize proportions of one or more combinations of the of predicted plurality of ingredients (114) through an AI-driven optimization technique to validate the nutritional consistency in the culinary item. In an embodiment, the AI-driven optimization techniques refer to the use of artificial intelligence processes to improve decision-making and enhance the efficiency of the system (100). The AI-driven optimization techniques are designed to find solution to complex problems.
[040] In an embodiment, the system (100) is configured to refine the optimization of the plurality of ingredients (114) based on feedback from the predictions of the ML model, whereby the ML model utilizes active learning for obtaining accuracy of the ML model. Moreover, the one or more processors (102) are configured to recommend a preferred culinary recipe by selecting a combination of ingredients from the predicted one or more combinations of ingredients with an optimized proportion of the plurality of ingredients (114) for obtaining the required nutritional value of the culinary item. In an embodiment, the system (100) predicts several possible combinations of ingredients that may be used in the preferred culinary recipe.
[041] In accordance with the present invention, Figure 2 depicts a typical network environment (200) for a system (100) for recommending culinary recipes (hereinafter referred to as “system”). In an embodiment of the invention, the network environment (200) is a public network environment which includes various servers and computing devices. In yet another embodiment of the invention, the network environment (200) is a private network with a limited number of computing devices such as personal computers, servers, laptops, mobile phones, etc.
[042] Referring yet to Figure 2, in an embodiment of the invention, the system (100) is implemented in one or more user devices (206). In the same embodiment, the user devices 106 include multiple applications such as web applications that may be running to perform several functions, as required by different users. In and embodiment, the system (100) is implemented in a computing device, such as a laptop computer, a desktop computer, a notebook, a workstation, a mainframe computer, a server, and the like. The user devices (206) may be implemented as, but are not limited to, desktop computers, hand-held devices, laptops or other portable computers, tablet computers, mobile phones, PDAs, Smartphones, land-line phones, and the like. In an embodiment, the user devices (206) are capable of providing content, through a network (204).
[043] In one aspect, the present invention discloses a system for recommending culinary formulation based on user requirements (112) and culinary information for providing at least one culinary recipe responsive to dynamically varying user response, learned user behavior, and user contextual information. Referring to Figure 2, in an embodiment of the invention, the network (204) is a wireless or a wired network, or a combination thereof. In yet another embodiment, the network (204) is a collection of individual networks, interconnected with each other and functioning as a single large network (e.g., the internet or an intranet). Examples of such individual networks include, but are not limited to, Global System for Mobile Communication (GSM) network, Universal Mobile Telecommunications System (UMTS) network, Personal Communications Service (PCS) network, Time Division Multiple Access (TDMA) network, Code Division Multiple Access (CDMA) network, Next Generation Network (NGN), Public Switched Telephone Network (PSTN), and Integrated Services Digital Network (ISDN).
[044] In another aspect, and as seen in Figure 3, the present invention discloses a method (300) for recommending culinary recipes based on user requirements (112) and user information. According to the invention, at step (302) the method (300) includes receiving, data for a primary culinary recipe (110) for a culinary item corresponding to user requirements (112). The data (208) is unorganized and is either statistical in nature, or has categorical information from texts/ strings, or is in a form of lists, or any other form or combination thereof. At step (304) the method (300) includes, determining distribution of the plurality of ingredients (114) for the primary culinary recipe (110) for the culinary item, whereby the distribution is dynamic across a plurality of culinary groups, inferring a distinct weightage associated with the plurality of culinary groups.
[045] As per the invention, at step (306) the method (300) includes calculating, statistical measures for a nutritional value of each ingredient with a weighted mean, standard deviation, and confidence intervals, to estimate a probability of predefined range of the plurality of ingredients (114). At step (308) the method (300) includes training (308), a machine learning (ML) model using the statistical measures for one or more nutritional values, whereby the trained ML model is configured to predict one or more recipes with one or more combinations of ingredients to achieve a required nutritional value. In an embodiment, the plurality of ingredients (114) may have characteristics such as, but are not limited to, starch concentration, pH levels, moisture levels, and hygroscopicity of the ingredients.
[046] According to the invention, at step (310) the method (300) includes mapping one or more nutritional values of each of the plurality of ingredients (114) form one or more predicted recipes with nutritional value of the primary culinary recipe (110) using an artificial intelligence (AI) driven feature selection technique. The method (300) also includes applying, clustering techniques to group a plurality of ingredients (114) with same nutritional and one or more sensory characteristics and identifying, outlier ingredients by using anomaly detection to identify an ingredient mismatched with the nutritional profile or sensory characteristics of the primary culinary recipe (110). In an embodiment, the sensory characteristics includes at least one of taste, smell, texture, appearance, and color. The sensory characteristics also includes sound, for example, foods that are crunchy or crispy (like chips, crackers, or fried foods) often create a satisfying, loud sound when bitten into, fizzing sounds in carbonated drinks (like soda or sparkling water) contribute to the sensory experience.
[047] At step (312) the method (300) includes recommending, a preferred culinary recipe by selecting a combination of ingredients from the predicted one or more combinations of ingredients with an optimized proportion of the plurality of ingredients (114) for obtaining the required nutritional value of the culinary item.
[048] Advantageously, the system and method for recommending culinary recipes of the present invention offers significant advantages by providing recommendations for culinary recipes that are specifically tailored to meet the comprehensive and dynamic requirements of individual users. The system leverages advanced algorithms, data analytics, and machine learning techniques to process user data and offer personalized food recipes that not only satisfy nutritional needs but also accommodate health conditions, personal preferences, and sustainability goals. The present invention also enables more effective and accurate dietary recommendations, promoting better health outcomes and enhancing the overall user experience in meal planning and food selection. Further, the present invention uses AI processes which can tailor recipe suggestions based on individual preferences, dietary restrictions, health goals, and taste preferences. Hence, the present invention ensures that users receive recommendations that are relevant to their specific needs and likes, increasing satisfaction and engagement. The AI techniques used in the present invention, helps to quickly suggest recipes with optimum nutritional values. The system combines topic modeling and knowledge graphs which is a powerful method to enhance text analysis and knowledge discovery. This approach integrates unstructured and structured data, making NLP applications more accurate, interpretable, and context-aware.
[049] While various aspects and embodiments have been disclosed hereinabove, other aspects and embodiments will be apparent to those skilled in the art. However, it will be apparent to those skilled in the art that various changes and modification may be made without departing from the scope of the invention as defined in the following claims.
, Claims:1. A system (100) for recommending culinary recipes based on user requirements (112) and user information, the system (100) comprising:
one or more processors (102); and
a memory (104) operatively coupled to one or more processors (102), wherein the memory (104) comprises a set of instructions which, when executed by the one or more processors (102), cause the one or more processors (102) to:
receive data (208) for a primary culinary recipe (110) of a culinary item corresponding to user requirements (112);
determine distribution of a plurality of ingredients (114) for the primary culinary recipe (110) of the culinary item, whereby the distribution is dynamic across a plurality of culinary groups, inferring a distinct weightage associated with the plurality of culinary groups;
calculate statistical measures for a nutritional value of each ingredient with a weighted mean, standard deviation, and confidence intervals, to estimate a probability of predefined range of the plurality of ingredients (114);
train a machine learning (ML) model using the statistical measures for one or more nutritional values, whereby the trained ML model is configured to predict one or more recipes with one or more combinations of ingredients to achieve a required nutritional value;
map one or more nutritional values of each of the plurality of culinary ingredients form one or more predicted recipes with nutritional value of the primary culinary recipe (110) using an artificial intelligence (AI) driven feature selection technique; and
recommend a preferred culinary recipe by selecting a combination of ingredients from the predicted one or more combinations of ingredients with an optimized proportion of the plurality of ingredients (114) for obtaining the required nutritional value of the culinary item.

2. The system (100) as claimed in claim 1, wherein one or more processors (102) is configured to analyze relationships between the predicted one or more combinations of ingredients for formulation the culinary item corresponding to the predefined range of the plurality of ingredients (114).

3. The system (100) as claimed in claim 2, wherein one or more processors (102) are configured to detect anomalies in the relationships for mapping one or more nutritional values of each of the plurality of culinary ingredients.

4. The system (100) as claimed in claim 1, wherein one or more processors (102) are configured to optimize proportions of one or more combinations of the of predicted plurality of ingredients (114) through an AI-driven optimization technique to validate the nutritional consistency in the culinary item.

5. The system (100) as claimed in claim 1, wherein one or more processors (102) are configured to refine the optimization of the plurality of ingredients (114) based on feedback from the predictions of the ML model, whereby the ML model utilizes active learning for obtaining accuracy of the ML model.

6. The system (100) as claimed in claim 1, wherein one or more processors (102) are configured to clean and preprocess the data to handle missing values, normalize nutritional values, and transform a categorical data corresponding to the plurality of ingredients (114) into usable numerical features.

7. The system (100) as claimed in claim 1, comprises one or more AI-driven distance matrices configured to analyze similarity between the plurality of ingredients (114) and ability to substitute with another ingredient while maintaining flavor, texture, and nutritional value of the culinary item.

8. The system (100) as claimed in claim 1, wherein the AI-driven feature selection technique is configured to identify at least one of the features including, taste, texture, and cost along with the nutritional value.

9. The system (100) as claimed in claim 1, wherein the received data comprises a plurality of primary culinary recipes with a plurality of recipes and a set of features responsible for the preferred culinary recipe.

10. The system (100) as claimed in claim 1, wherein the plurality of instructions causes one or more processors (102) to analyze relationship between the predicted one or more combinations of ingredients by constructing and utilizing the one or more AI-driven distance matrices to quantify clustering of the plurality of culinary groups.

11. The system (100) as claimed in claim 1, wherein the plurality of ingredients (114) with nutritional data is extracted from a database by using a combination of Topic Modeling and Knowledge Graphs (KGs) technique, thereby ensuring the extracted data is structured, whereby the database (116) is hosted on a server (118).

12. The system (100) as claimed in claim 1, wherein the one or more processors (102) are configured to correlate the predicted features with the nutritional requirements, sensory requirements, and functional requirements of the primary culinary recipe.

13. The system (100) as claimed in claim 1, wherein the weighted mean is based on average of the plurality of culinary groups sharing a same type of the culinary item.

14. A method (300) for recommending culinary recipes based on user requirements (112) comprising:
receiving (302), data for a primary culinary recipe (110) for a culinary item corresponding to user requirements (112);
determining (304), distribution of the plurality of ingredients (114) for the primary culinary recipe (110) for the culinary item, whereby the distribution is dynamic across a plurality of culinary groups, inferring a distinct weightage associated with the plurality of culinary groups;
calculating (306), statistical measures for a nutritional value of each ingredient with a weighted mean, standard deviation, and confidence intervals, to estimate a probability of predefined range of the plurality of ingredients (114);
training (308), a machine learning (ML) model using the statistical measures for one or more nutritional values, whereby the trained ML model is configured to predict one or more recipes with one or more combinations of ingredients to achieve a required nutritional value;
mapping (310), one or more nutritional values of each of the plurality of ingredients (114) form one or more predicted recipes with nutritional value of the primary culinary recipe (110) using an artificial intelligence (AI) driven feature selection technique; and
recommending (316), a preferred culinary recipe by selecting a combination of ingredients from the predicted one or more combinations of ingredients with an optimized proportion of the plurality of ingredients (114) for obtaining the required nutritional value of the culinary item.

15. The method (300) as claimed in claim 14, comprises:
applying (312), clustering techniques to group a plurality of ingredients (114) with same nutritional and one or more sensory characteristics; and
identifying (314), outlier ingredients by using anomaly detection to identify an ingredient mismatched with the nutritional profile or sensory characteristics of the primary culinary recipe (110).

16. The method (300) as claimed in claim 15, wherein the sensory characteristics comprises at least one of taste, smell, texture, appearance, and color.