DESC:RESERVATION OF RIGHTS
A portion of the disclosure of this patent document contains material which is subject to intellectual property rights such as, but are not limited to, copyright, design, trademark, IC layout design, and/or trade dress protection, belonging to Jio Platforms Limited (JPL) or its affiliates (hereinafter referred as owner). The owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent files or records, but otherwise reserves all rights whatsoever. All rights to such intellectual property are fully reserved by the owner.

FIELD OF INVENTION
[0001] The present disclosure relates to the field of microlending services. More particularly the present disclosure relates to system and method for edge microlending services using a machine learning model on a distributed platform.

BACKGROUND OF THE INVENTION
[0002] The following description of related art is intended to provide background information pertaining to the field of the disclosure. This section may include certain aspects of the art that may be related to various features of the present disclosure. However, it should be appreciated that this section be used only to enhance the understanding of the reader with respect to the present disclosure, and not as admissions of prior art.
[0003] Microlending is a process of lending money in small amount unlike loans from various banks. Microlending is a usually performed by individual entities unlike the huge financial institutions. Such microlending process helps many individuals specially for edge or remote communities in supporting such communities by providing financial aid. In a conventional approach, the microlending or any such financial aid are provided by physically knowing one or more users of any such communities, and by physically verifying the associated documents in order to provide the financial assistance. However, due to the human intervention and lack of documentation and authentication of the documents
[0004] In comparison with the conventional approach, a newer approach is used to validate the details of the user on a platform upon feeding the user details and verifying the same by comparison. However, due to the human intervention the validation and processing of data is not very accurate. Also, understanding the requirements of the user and matching the data to validate the authenticity is mostly done manually. Due to such limitations, both the conventional approach and the newer approaches are less reliable and less efficient.
[0005] There is, therefore, a need of an improved system and method for edge microlending services to address the aforementioned issues(s).

OBJECTS OF THE PRESENT DISCLOSURE
[0006] Some of the objects of the present disclosure, which at least one embodiment herein satisfies are as listed herein below.
[0007] It is an object of the present disclosure to provide for a system and a method for leveraging a combination of emerging technologies such as microservices, blockchain technology, and machine learning at an edge community to explore a microfinancing services application at the edge community.
[0008] It is an object of the present disclosure to provide for developing an edge model that could be hosted in the cloud or at an edge-server (if available) to cater to the needs of an edge community.
[0009] It is an object of the present disclosure to provide for including many edge communities and provide divergence in microlending services.
[0010] It is an object of the present disclosure to provide for a system and a method for utilizing edge learning with accurate predictions based on the nature of an edge community.
[0011] It is an object of the present disclosure to provide for a system and a method for leveraging other sustainable edge applications as well.

SUMMARY
[0012] This section is provided to introduce certain objects and aspects of the present invention in a simplified form that are further described below in the detailed description. This summary is not intended to identify the key features or the scope of the claimed subject matter.
[0013] In order to achieve the aforementioned objectives, the present invention provides a system and method for assisting and managing microlending service. The system may include one or more processors coupled with a memory wherein the memory may store instructions which when executed by the one or more processors causes the system to: receive, a set of data packets pertaining to information associated to a user involved in one of lending or borrowing funds, the set of data packets may be received from a plurality of databases communicatively coupled to one or more servers; extract, by a feature extraction engine, a first set of features from the set of data packets received, the first set of features pertaining to a plurality of predefined parameters associated with a loan status of the user; determine, by an ML engine, a microlending model based on a predefined set of instructions and the first set of features extracted; train, by the ML engine, the microlending model based on the predefined set of instructions, the first set of features extracted; and based on the trained microlending model, predict, by the ML engine, a probability for the user to be eligible to borrow or lend a predicted amount of funds.
[0014] In another aspect, the present disclosure includes a method for assisting and managing microlending service. The method may be executed by a processor, and includes the steps of: receiving, by a processor, a set of data packets pertaining to information associated to a user involved in one of lending or borrowing funds, the set of data packets are received from a plurality of databases communicatively coupled to one or more servers; extracting, by a feature extraction engine, a first set of features from the set of data packets received, the first set of features pertaining to a plurality of predefined parameters associated with a loan status of the user. Further, the method may include the step of determining, by an ML engine, a microlending model based on a predefined set of instructions and the first set of features extracted. Furthermore, the method may include the step of training, by the ML engine, the microlending model based on the predefined set of instructions, the first set of features extracted. Based on the trained microlending model, the method may also include the step of predicting, by the ML engine, a probability for the user to be eligible to borrow or lend a predicted amount of funds.

BRIEF DESCRIPTION OF DRAWINGS
[0015] The accompanying drawings, which are incorporated herein, and constitute a part of this invention, illustrate exemplary embodiments of the disclosed methods and systems in which like reference numerals refer to the same parts throughout the different drawings. Components in the drawings are not necessarily to scale, emphasis instead being placed upon clearly illustrating the principles of the present invention. Some drawings may indicate the components using block diagrams and may not represent the internal circuitry of each component. It will be appreciated by those skilled in the art that invention of such drawings includes the invention of electrical components, electronic components or circuitry commonly used to implement such components.
[0016] FIG. 1A illustrates an exemplary system architecture of the proposed system for assisting microlending service, in accordance with an embodiment of the present invention;
[0017] FIG. 1B illustrates a block diagram representing of a system to assist a microlending service at an edge platform and a cloud platform, in accordance with an embodiment of the present disclosure;
[0018] FIG. 2A illustrates an exemplary representation of proposed system /centralized server for accessing content stored in a network, in accordance with an embodiment of the present disclosure;
[0019] FIG. 2B illustrates an architecture representing microservices based platform for the system for edge microlending service of FIG. 1, in accordance with an embodiment of the present disclosure;
[0020] FIG. 2C illustrates an exemplary representation of the proposed method in accordance with an embodiment of the present disclosure;
[0021] FIG. 3 illustrates an exemplary embodiment of a block diagram representing different methods used in the system to the assist microlending service at the edge platform and the cloud platform of FIG. 1, in accordance with an embodiment of the present disclosure;
[0022] FIG. 4A illustrates an exemplary embodiment of a block diagram representing a process involved in generating a rate for corresponding users on the edge platform and the cloud platform of FIG. 1, in accordance with an embodiment of the present disclosure;
[0023] FIG. 4B illustrates an exemplary embodiment of a block diagram representing a process involved in generating a decision associated to the microlending service on a distributed platform of FIG. 1, in accordance with an embodiment of the present disclosure;
[0024] FIG. 5 illustrates an exemplary embodiment of a block diagram representing an architecture of the system to assist microlending service of FIG. 1, in accordance with an embodiment of the present disclosure; and
[0025] FIG. 6 illustrates an exemplary computer system in which or with which embodiments of the present invention can be utilized in accordance with embodiments of the present disclosure.
[0026] The foregoing shall be more apparent from the following more detailed description of the invention.

BRIEF DESCRIPTION OF INVENTION
[0027] In the following description, for the purposes of explanation, various specific details are set forth in order to provide a thorough understanding of embodiments of the present disclosure. It will be apparent, however, that embodiments of the present disclosure may be practiced without these specific details. Several features described hereafter can each be used independently of one another or with any combination of other features. An individual feature may not address all of the problems discussed above or might address only some of the problems discussed above. Some of the problems discussed above might not be fully addressed by any of the features described herein.
[0028] The present disclosure relates to the field of microlending services. More particularly the present disclosure relates to system and method for an assisting microlending services using a machine learning model on a distributed platform.
[0029] Referring to FIG. 1A that illustrates an exemplary network architecture (100) in which or with which system 110 of the present disclosure can be implemented, in accordance with an embodiment of the present disclosure. As illustrated, the exemplary architecture (100) includes a system 110 equipped with an machine learning (ML) engine (216) for assisting in microlending services to a plurality of users 102-1, 102-2,…. 102-n (hereinafter interchangeably referred as user or lender or borrower; and collectively referred to as users 102). Each user may be associated with at least one computing device 104-1, 104-2,…. 104-n (hereinafter interchangeably referred as a smart computing device; and collectively referred to as 104). The users 102 may interact with the system 110 by using their respective computing device 104, wherein the computing device 104 and the system 110 may communicate with each other over a network 106. The system 110 may be associated with a centralized server 112 that may be one or more servers comprising one or more edge servers 112-1, one or more cloud servers 112-2 or a combination thereof but not limited to the like.
[0030] Examples of the computing devices 104 can include, but are not limited to, a computing device 104 associated with banking and fund lending or borrowing based firms, a smart phone, a portable computer, a personal digital assistant, a handheld phone and the like.
[0031] Further, the network 106 can be a wireless network, a wired network, a cloud or a combination thereof that can be implemented as one of the different types of networks, such as Intranet, BLUETOOTH, MQTT Broker cloud, Local Area Network (LAN), Wide Area Network (WAN), Internet, and the like. Further, the network 106 can either be a dedicated network or a shared network. The shared network can represent an association of the different types of networks that can use variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol/Internet Protocol (TCP/IP), Wireless Application Protocol (WAP), and the like. In an exemplary embodiment, the network 104 can be anHC-05 Bluetooth module which is an easy to use Bluetooth SPP (Serial Port Protocol) module, designed for transparent wireless serial connection setup.
[0032] According to various embodiments of the present disclosure, the system 110 can provide for machine learning (referred to as ML hereinafter), deep learning (referred to as DL hereinafter), concepts of neural network techniques. The technique and other microlending model involved in the use of the technique can be accessed from a plurality of databases in the server.
[0033] In an aspect, the system 110 can receive, a set of data packets pertaining to information associated to a user involved in one of lending or borrowing funds, the set of data packets maybe received from the plurality of databases communicatively coupled to one or more servers; extract, by a feature extraction engine 214 (Ref. FIG. 2A), a first set of features from the set of data packets received, said first set of features pertaining to a plurality of predefined parameters associated with a loan status of the user 102. The system 110 may further determine, by the ML engine 216, the microlending model based on a predefined set of instructions and the first set of features extracted. The ML engine 216 may then train the microlending model based on the predefined set of instructions, the first set of features extracted; and based on the trained microlending model, the ML engine 216 may predict a probability for the user to be eligible to borrow or lend a predicted amount of funds. For example, the prediction of whether a loan will default or not is based on the various decision enabling parameters and on any available information about the borrowers’ past transaction as well as aggregate metrics related to the community that the borrower resides in.
[0034] In an embodiment, the plurality of predefined parameters comprise a default value, a charged off value, a fully paid value, a current value, an issued value, a grace period value, a first late value, a second late value or a combination thereof. In one exemplary embodiment, the raw data may be retrieved form one or more database including the plurality of pre-defined parameters, one such parameter may include a loan status of the borrower, wherein the loan status may be one of the default value, the charged off value, the fully paid value, the current value, the issued value, the grace period value, the first late value, the second late value or the combination thereof.
- Wherein, for example the Default value may be defined as a value when the borrower was unable to repay one or more previous loans and hence the loan was defaulted. Value = 1;
- The Charged – off may be defined as a value when there is no reasonable expectation of further payment by the borrower. Value = 2;
- The Fully paid value may be defined as the amount of the loan taken previously by the borrower which has been fully repaid. Value = 5;
- The Current value may be defined as the loan amount which is up to date on all outstanding payments. Value = 0
- The Issued value may be defined as the loan amount which has passed all initial checks and has been funded by the lender. Value = 0
- The Grace Period may be defined as the Loan amount which may have past due date and within the 15-day grace period (say for example). Value = 4
- The first Late value (for example 16-30 days) may be defied as the Loan amount which has not been current for 16 to 30 days. Value = 3
- The second Late value (for example 31-120 days) may be defined as the Loan amount which has not been current for 31 to 120 days. Value = 2.
[0035] In an embodiment, the set of data packets data (interchangeably referred to as raw data herein) may be subjected to feature engineering by the feature extraction engine to extract the first set of features having a high positive correlation with a target variable, wherein the target variable may be associated with the probability for the user to be eligible to borrow or lend the predicted amount of funds. For example, the feature extraction engine may enable building of an unbiased predictor. Using feature extraction engine, relevant features may be selected which have a high positive correlation with the target variable, thus increasing the predictive power of the system
[0036] In an embodiment, one or more target variables are converted into numeric values by categorizing and assigning weights to the target values defined based on the loan status.
[0037] In an embodiment, the first set of features may be pre-processed to obtain a second set of features to compute a feature importance value for each of the second set of features. In an embodiment, a cross validation module may be applied on the first set of features in order to split and validate the set of data packets to obtain an adequate set of classes associated with the first set of features. For example, initially, target variables which may be a string feature may be converted into numeric values by categorizing and assigning weights to the values, which may be defined based on the loan status. Further, the one or more features of the raw data may be pre-processed to obtain a standard set of features. Further, Extra Trees Classifier may be applied on the set of obtained standard features to compute a feature importance value for each of the standard set of features. Subsequently, in order to split and validate the data, a K-Fold Cross Validation technique may be applied on the set of features to obtain an adequate set of classes associated to the set of features.
[0038] In an embodiment, a multiservice platform may be operatively coupled to the processor to assist and manage microlending of a plurality of users across a plurality of distributed computing user devices 104 associated with a blockchain. In an embodiment, the multiservice platform further comprises a lender interaction module, a borrower interaction module, one or more prediction models, Inter Planetary File System 110 (IPFS) front end module, a blockchain front end module, the edge server and the cloud server. The multiservice platform may be operatively coupled to the plurality of databases.
[0039] Consequently, details associated to one of the one or more users is retrieved from one or more sources to predict a probability for the borrower to be eligible to borrow a sum of amount or funds from the lender.
[0040] FIG. 1B illustrates a block diagram representing of the system 110 to an assisting microlending service at an edge platform and a cloud platform, in accordance with an embodiment of the present disclosure. The system 110 includes an edge module 152 and a cloud module 154. On both the edge module 152 and the cloud module 154, initially a set of raw data may be extracted in order to train a data model in step 156, using one of a machine learning technique, an artificial intelligence technique, or the like. As used herein, the term ‘Machine learning technique’ is defined as a study of computer algorithms that improve automatically through experience and by the use of data. It is seen as a part of artificial intelligence. Also, the term ‘artificial intelligence’ is defined as an intelligence demonstrated by machines, unlike the natural intelligence displayed by humans and animals, which involves consciousness and emotionality. Further, upon training the data model, the raw data is preprocessed and one or more features are extracted from the data model in step 158. On extracting the one or more features, the data models are trained using the one or more extracted features in step 160. In one exemplary embodiment, the data model may be trained using Millions of sample raw data to enable the data model to understand the kind of data required and to generate more accurate output. In one exemplary embodiment, the raw data may be representative of the details associated to one or more user involved in one of lending or borrowing funds. In such embodiment, the one or more user may include one of a lender, a borrower, or a combination thereof in step 162 and 164.
[0041] FIG. 2A illustrates an exemplary representation of proposed system /centralized server for accessing content stored in a network, in accordance with an embodiment of the present disclosure. In an aspect, the system 110 /centralized server 112 may comprise one or more processor(s) 202. The one or more processor(s) 202 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 204 of the system 110. 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 RAM, or non-volatile memory such as EPROM, flash memory, and the like.
[0042] In an embodiment, the system 110/ centralized server 112 may include an interface(s) 206. The interface(s) 206 may comprise a variety of interfaces, for example, interfaces for data input and output devices, referred to as I/O devices, storage devices, and the like. The interface(s) 206 may facilitate communication of the system (102). The interface(s) 206 may also provide a communication pathway for one or more components of the centralized server 112. Examples of such components include, but are not limited to, processing engine(s) 208 and a database 210.
[0043] The processing engine(s) 208 may be implemented as a combination of hardware and programming (for example, programmable instructions) to implement one or more functionalities of the processing engine(s) 208. In examples described herein, such combinations of hardware and programming may be implemented in several different ways. For example, the programming for the processing engine(s) 208 may be processor executable instructions stored on a non-transitory machine-readable storage medium and the hardware for the processing engine(s) 208 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 processing engine(s) 208. In such examples, the system 110 /centralized server 112 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 system 110 /centralized server 112 and the processing resource. In other examples, the processing engine(s) 208 may be implemented by electronic circuitry. The processing engine 208 may include one or more engines selected from any of a data acquisition engine 212, feature extraction engine 214, ML engine 216, and other units 218.
[0044] The data acquisition engine 212 may be configured to receive a set of data packets pertaining to information associated to a user involved in one of lending or borrowing funds, the set of data packets maybe received from the plurality of databases communicatively coupled to one or more servers.
[0045] The feature extraction engine may be configured to extract a first set of features from the set of data packets received, said first set of features pertaining to a plurality of predefined parameters associated with a loan status of the user.
[0046] The ML engine 216 may determine a microlending model based on a predefined set of instructions and the first set of features extracted. The ML engine 216 may then train the microlending model based on the predefined set of instructions, the first set of features extracted; and based on the trained microlending model, the ML engine 216 may predict a probability for the user to be eligible to borrow or lend a predicted amount of funds.
[0047] FIG. 2B illustrates an architecture representing microservices based platform for the system 200 for the assisting microlending service of FIG. 1B, in accordance with an embodiment of the present disclosure. An interface 222 which is associated to the one or more users is operatively coupled to a web server 224, which is associated to a multiservice platform 226. The multiservice platform 226 includes a lender interaction module 228, a borrower interaction module 230, one or more prediction models 232, Inter Planetary File System (IPFS) front end module 234, a blockchain front end module 236, the remote edge server 238 and the remote cloud server 240. The multiservice platform 206 is operatively coupled to a plurality of databases such as an NOSQL or an SQL database 242, IPFS database 244, a blockchain database 246, a remote cloud server database 248-1and a remote edge server database 248-2. In an embodiment, the lender interaction module 228 may be configured to execute a set of instructions though which a lender can interact with one or more users about one or more loan details. In an embodiment, the borrower interaction module may be configured to execute a set of instructions through which a borrower can interact with a lender about the one or more loan details. Further, the one or more prediction models can predict whether a loan will default or not based on decision enabling parameters extracted from the lender interaction module 228 and on any available information about the borrowers’ past transaction as well as one or more aggregate metrics related to a community that the borrower resides in extracted from the borrower interaction module 230. In one embodiment, the lender interaction (LI) module 228, the borrower interaction (BI) module 230 may execute an inference model (IM) for prediction of the IM and store information in a database (SDB) or an inter planetary file system (IPFS) 234 and may further interact with the blockchain front end module 236 to execute a smart contract microservice for determine whether to : lend (BSC), interact with a blockchain platform to record information about a lending transaction in a distributed ledger (BREC), interact with the remote edge server 238 for an edge service (ES) such as to perform inferencing with an edge ML model or ensembled model across ML models, interacting with the remote cloud server 240 for a cloud service (CS) such as to perform batch learning of an edge model in the cloud, or the like. Thus, the microservices can also be chained to create a service-chain that executes a sequence of tasks such as BI -> LI -> IM -> BSC -> BREC -> LI -> BI to process the microlending request from a borrower, and provide the outcome of processing back to the borrower.
[0048] In an embodiment, the prediction can be directly based on the borrower’s loan detail such as credit grading assigned based on the current loan parameters using Machine Learning and based on the borrower’s past transactions for calculating credit grading. The one or more prediction modules can include a Logistic Regression, a Light Gradient Boosting Framework (LGBM) a Neural Network and not limited to the mentioned ML models and can be extended to support any other ML model. The Inter Planetary File System (IPFS) front end module 234 may be configured for storing and sharing data in the system. The IPFS front end module 234 may use content-addressing to uniquely identify each file in a global namespace connecting all computing devices. Furthermore, IPFS is used to store large amounts of data off-chain and place the immutable, permanent IPFS links as well as the cryptographic hash of our prediction model into the Quorum network. The smart contract of the blockchain network basically implements three main functionalities such as request IPFS, send IPFS and retrieve IPFS. The request IPFS stores the address of the calling node which is then served one by one by owner of the model using the send IPFS method of the smart contract. Send IPFS then maps the generated IPFS hash, on to the address retrieved from the request IPFS. The user executing the request IPFS functionality can then check 'retrieve IPFS' to see if the function has received the cryptographic hash on its address, and if yes then retrieve the function. Once the transaction is complete, the hash returned is then recorded in the network. This way a secure base for the service which cannot be tampered may be provided.
FIG. 2C illustrates an exemplary representation of the proposed method 250 in accordance with an embodiment of the present disclosure. As illustrated, in an aspect the propose method 250 for assisting and managing microlending service. The method 250 may be executed by a processor, and includes at 252, the step of: receiving, by the processor, a set of data packets pertaining to information associated to a user involved in one of lending or borrowing funds, the set of data packets may be received from a plurality of databases communicatively coupled to one or more servers; and at 254, the step of extracting, by a feature extraction engine, a first set of features from the set of data packets received, the first set of features pertaining to a plurality of predefined parameters associated with a loan status of the user. Further, the method 250 may include at 256, the step of determining, by an ML engine, a microlending model based on a predefined set of instructions and the first set of features extracted. Furthermore, the method 250 may include at 258 the step of training, by the ML engine, the microlending model based on the predefined set of instructions, the first set of features extracted. Based on the trained microlending model, the method 250 may also include at 260, the step of predicting, by the ML engine, a probability for the user to be eligible to borrow or lend a predicted amount of funds.
[0049] Turning to FIG. 3, FIG. 3 illustrates an exemplary embodiment of a block diagram representing different methods used in the system 300 to the assisting microlending service at the edge platform and the cloud platform of FIG. 1, in accordance with an embodiment of the present disclosure. One or more parameters associated to microlending service may be used as inputs 302 and the ML technique may be applied on the said input using the ML model. In one embodiment, the ML may be one of a Logistic Regression model 304, a Light GBM model 306, a Neural Network model 308 or a combination thereof, which may be used in predicting the probability of a loan defaulting. All the ML models may be used to generate an ensemble model 310 which may be configured to compute the prediction.
[0050] Further, in case of the logistic regression model 304, the logic transformed prediction probability may be a linear function of the target variable values. The input data (X) may be combined linearly using the coefficient values to predict a target variable (Y) which loan status = default (for example). The logistic regression model 304 uses C = 0.0001, where C is the Inverse of Regularization, Strength, Tolerance = 1e 4, with a Lib Linear Solver and Class Weight = none.
[0051] Furthermore, in case of the neural network model 308, basic idea behind a neural network is to simulate a lot of densely connected neurons (brain cells) artificially so that one can get it to learn things, draw inferences, make patterns and take an informed decision in a humanlike way. The neural network has a single input layer may include a plurality of neurons corresponding to the number of independent variables in the dataset. input dimensions as that of the number of input layers and activation function = sigmoid, whereas our output layer is also single layers consisting of 1 neuron with softmax activation function wherein there is one node per class label. Finally, there are two hidden layers of made up of a pre-defined number of neurons each. In such scenarios, a general consensus to keep the size of the hidden layer in between the size of the output layer and the input layer, with the sigmoid activation function.
[0052] In case of the Light GBM (LGBM) model 306, LGBM is known for its higher efficiency, less usage memory, higher accuracy as well as faster training speed. The LGBM model may use N estimator = 1000, learning rate = 0:02, max depth = 8, reg alpha = 0:04, n threads = 4, min child weight = 40 and The resulting output of all the models is a probability between 0 to 1, of the loan defaulting which is then converted to a rating from 0-5 by using an equation:
[0053] Rating = Round ((1 - Probability) * 5). If the probability of the loan defaulting is 0.8 then the rating of the borrower loan will be 1. Similarly, if the chances of the loan defaulting are 0.2, then the rating of the borrower loan will be 4.
[0054] Furthermore, for, another approach may be used by calculating the rating for the borrower, a mathematical model may be used which computes an amount repaid and the rate of interest of a particular loan the most importance based on the loan status. The model normalizes the summation of all the loan entries of a particular borrower to get a value between 0 and 5. The summation of rating for all the individual loan may be defined using:
[0055] R = ? (W * (A Repaid/ A Repaid + A Outstanding) + IR/ MIR), Where W is the weight of the loan, IR is the Interest Rate of the individual loan, MIR is the mean of all the interest rates in the data set, a Repaid is the amount paid back in full at the time of credit rating calculation and an Outstanding is the total loan amount minus the Amount Repaid. Where, the Aggregated and Normalized Rating for the borrower is:
[0056] Final Rating = Round ((?Rating/ 6 *N) * 5), Where N is the number of valid loan entries and Round function rounds the value to the nearest integer.
[0057] In yet another approach, both the approaches mentioned above. Both the approaches are assigned weights according to their importance in calculating the credit of the borrower. This way the lenders have a complete picture of the borrowers and their capability of repaying the loan listed by them. The lender can then weigh the risks and benefit to make a decision regarding the loan funding. In this approach, dynamic weights to the borrower's past transaction and its effect on the rating may be assigned. The borrower's current loan parameters and its likeness are given a slightly lesser weightage. Here, the weight assigned to both approaches depends on the borrower's past transactions and their capability. If the borrower is a frequent user, more weightage was given to the past transactions as it is best reflective of his/her repaying capabilities. Based on the number of past transactions up to a threshold _ = 100, weights are assigned, above which past transactions are assigned a weight of 0.9. Once rating is obtained by using both the approaches, we combine then using the following formula: Rating (R) = W1 * R1 +W2 * R2, Where R1 and R2 are the ratings from the first and the second approach respectively.
[0058] A grading based on a generic anonymized model which includes of all the transactions in the network, stored in the cloud can be combined with a grading based on past transactions for a specific borrower at the edge to determine an overall grading for the borrower at the edge. Thus, distributed processing with combined execution of microservices at the cloud and the edge could be utilized to determine an overall grading for a borrower in a community of users being serviced at the edge. It should be noted that one could use a value of modified version a of given by a’=?*a where ? [0,1], so that some weightage to the grading obtained from the cloud is always utilized (when ? ? [0,1)) even if a = 1 in the weighted grade estimation. In this case g = (1 - a) * g1 + a * g2. In one embodiment, the microservices may include at least one of lender interaction (LI), borrower interaction (BI), executing an inference model (IM) for prediction (IM), storing information in a database (SDB) or storing in a file system (SFS), interacting with a blockchain platform to execute a smart contract microservice for determine whether to lend (BSC), interacting with a blockchain platform to record information about a lending transaction in a distributed ledger (BREC), interacting with an edge server for an edge service (ES) such as to perform inferencing with an edge ML model or ensembled model across ML models, interacting with a cloud server for a cloud service (CS) such as to perform batch learning of an edge model in the cloud, or the like. These microservices may interact with requisite platforms such as a database (SQL/noSQL), file storage platform (such as IPFS), a blockchain platform, an edge server, a cloud server, or the like. to accomplish different tasks. Microservices can also be chained to create a service-chain that executes a sequence of tasks such as BI -> LI -> IM -> BSC -> BREC -> LI -> BI to process the microlending request from a borrower, and provide the outcome of processing back to the borrower.
[0059] Furthermore, IPFS is used to store large amounts of data off-chain and place the immutable, permanent IPFS links as well as the cryptographic hash of our prediction model into the Quorum network. The smart contract of the blockchain network basically implements three main functionalities such as request IPFS, send IPFS and retrieve IPFS. The request IPFS stores the address of the calling node which is then served one by one by owner of the model using the send IPFS method of the smart contract. Send IPFS then maps the generated IPFS hash, on to the address retrieved from the request IPFS. The user executing the request IPFS functionality can then check 'retrieve IPFS' to see if the function has received the cryptographic hash on its address, and if yes then retrieve the function. Once the transaction is complete, the hash returned is then recorded in the network. This way a secure base for the service which cannot be tampered may be provided.
[0060] Turning to FIGs. 4A and 4B, FIG. 4A illustrates an exemplary embodiment of a block diagram representing a process 400 involved in generating a rate for corresponding users on the edge platform and the cloud platform of FIG. 1B, in accordance with an embodiment of the present disclosure. FIG. 4B illustrates an exemplary embodiment of a block diagram representing a process involved in generating a decision associated to the microlending service on a distributed platform of FIG. 1B, in accordance with an embodiment of the present disclosure. Initially a lender 402 upon completing a validation process, registers on the blockchain platform in step 404, simultaneously, a borrower 406 upon completing a validation process, registers on the blockchain platform in step 408. Consequently, on completing the validation process, based on the requirements for the microlending service, the system 400 generates a rating based the borrower’s transaction and another rating based on loan parameters in step 410, 412 and 414. On combining both the rating values, an overall weighted decision is generated by the system in step 416. Simultaneously, the rating based on loan parameters (as in step 414), the same is displayed on the cloud model and the edge model respectively in step 418 and 420.
[0061] Furthermore, the prediction from the ML models is retrieved in step 422, and are transmitted to the blockchain platform along with the borrower data, community specific incentives, pre-defined policies and regulations in step 424, 426 and 428. Upon combining all the data and predictions, the blockchain platform upon using the VDLT smart contract in step 430, computes a decision associated to the microlending for the corresponding borrower in step 432.
[0062] FIG. 5 illustrates an exemplary embodiment of a block diagram representing an architecture 500 of the system to the assisting microlending service of FIG. 1B, in accordance with an embodiment of the present disclosure. The architecture 500 includes a client platform 502 to an API gateway 504 either directly or via an identity provider 506. The API gateway 504 is communicatively coupled to a multiservice platform which includes at least one of borrower interaction module 508, a lender interaction module 510, Inter Planetary File System (IPFS) front end module 512, one or more prediction models 514, a blockchain front end module 516, the remote edge server 518 and the remote cloud server 520. The multiservice platform is operatively coupled to one or more databases such as NOSQL or an SQL database 522, IPFS database 524, a blockchain database 526, a remote edge server database 528 and a remote cloud server database 530. Also, the static content 532 and the CND 534 are operatively coupled to the client platform 502.
[0063] In operation, to build an edge model, training data is taken from information from past microlending transactions to create a model, and to also iteratively refine an existing model based on any new training data. Training data can be supplied in batches of recent lending transactions to create or refine an ML model for microlending. Training for an edge model can be performed either at an edge data center or at a cloud data center. If training is performed in the cloud, then the ML model for microlending is realized in a cloud data center. Subsequently, the trained model can be transferred to an edge data center for inferencing to process new microlending applications.
[0064] Further, the ML model data can be pre-processed (such as eliminating spurious data rows that have inadequate information) and key features extracted (such as to remove fields that may not be of interest such as the address of a lender etc. and only minimal required information retained for training). Additionally feature-extraction could be performed using initial layers of a deep neural network that can automatically extract features from the input training data. The input feature data (along with output information such as whether a loan was granted by a lender to a specific borrower, or whether a loan was defaulted by the borrower or whether the money was indeed returned and the loan successfully completed by the borrower) are used for training in the network to create an ML model for microlending. Different ML techniques such as using random forest techniques, or traditional regression-based techniques, or gradient boosting techniques, or deep learning, or support vector machine techniques, or other ML techniques can be used. Ensembled-ML across ML models can also be utilized for more robust performance of the overall prediction system. Ensembled learning (as shown in FIG. 3) across different ML techniques can also be performed in the cloud, with an edge inference model transferred to an edge cloud and utilized at the edge cloud to enable loan processing at the edge. Alternatively, edge learning can be performed at the edge and utilized at the edge.
[0065] In one exemplary embodiment, different edge models can be created based on different demographic information such as a geographic region, or age-groups, income groups, or gender, or based on context such as whether the loan is for a real-estate purchase or for agriculture, or for education of a child, or based on other economic and social criteria, so that the decision on whether to grant a loan can be different based on different edge models that can emerge based on these different criteria or contexts, or the like.
[0066] Also, a cloud model may be created that aggregates different edge models across different regions with aggregation performed for the same age-groups, income-groups, or gender, or based on contexts across different edge regions. This helps in learning across different edge populations that can learn more general behavior rather than edge-specific behavior, and this could also help to reduce noise that may exist in the edge model. A cloud model could also aggregate edge models across different contexts or demographics criteria. In addition, a cloud model could be used as a starting model for the edge, to be utilized for microlending decisions for a new edge population for which an edge microlending service needs to be deployed.
[0067] Moreover, both a cloud model and an edge model can be utilized (as shown in FIG. 4A) to take a decision at the edge, with a weighted combining across the two models utilized to take a decision at the edge. In early phases of edge learning, the system 100 can give a greater weight to the decision arising from a cloud model relative to the edge model. Over time, as the edge model gets refined further, then the edge model can be accorded a greater weight relative to the cloud model to take an overall decision in the system 100.
[0068] Consequently, after a result is obtained based on an appropriate inferencing model for loan prediction for a specific borrower, additional processing (as shown in FIG. 4b) could be performed that may include additional criteria such as policy and regulatory constraints for the given context and input loan criteria and demographics criteria, and any constraints that a lender may have. Such additional processing may be performed by executing and invoking one or more a smart contract on the blockchain platform that process constraints to make a final determination regarding whether to grant a loan.
[0069] For example, government policies and regulations may provide special incentives for people of a certain income group category, a certain gender, a specific context, or the like. Similarly, lenders may have such preferences in policies to grant loans. Such constraints could be processed in a smart contract to make a final determination on whether or not to grant a loan to the borrower. Information regarding past loan grants for the same borrower can also be utilized to take a more effective decision in the system 100. Decisions taken for past loan requests, information about successful completion of a loan, or failed completion with a loan default, or the like, may be recorded in a blockchain ledger. Such information in the blockchain ledger provides for trusted immutable transparent information sharing across lenders, to help with trusted decision making in the system based on such trusted past recorded information in the system.
[0070] Furthermore, Different microservices can be used for different aspects of processing in the microlending system. The microservices can interact with requisite platforms such as a database (SQL or noSQL), file storage platform (such as IPFS), a blockchain platform, the edge server, the cloud server, or the like, to accomplish different tasks. Microservices can also be chained to create a service-chain that executes a sequence of tasks to process a loan request from a borrower, and provide the outcome of processing back to the borrower.
[0071] Further, in a situation if the microlending process is successful, then such a loan could be granted on a Traceable Secure Digital Currency Blockchain (TSDCB) platform and the money lent on such a platform utilizing the digital or virtual currency supported by such a platform, and utilizing traceable identities associated with both the borrower and the lender. The borrower can then leverage the TSDCB platform to utilize the available digital currency or virtual currency balance in the user’s account to addressing the borrowing need for which the loan was granted. The borrower may also receive payments from the borrower’s employer, where a portion of the earned salary can be used to pay off the loan using the same digital currency or the virtual currency. The digital currency or the virtual currency may have to be converted to an alternate currency for some of such transactions. Over a period of time, the TSCDB platform could help in full repayment of the loan to the lender, with provenance of information related to transactions processed on the platform for the borrower. Such provenance information can also be used to determine a reputation score for the borrower, the lender, or the combination thereof in the system.
[0072] Various embodiments of the present disclosure enable the system to predict the probability for the lender and the borrower in the process of microlending service. Also, since the system uses the distributed platform, the system is highly secure and tamper proof. Since there is no human intervention, the system is more accurate henceforth more efficient and reliable.
[0073] FIG. 6 illustrates an exemplary computer system in which or with which embodiments of the present invention can be utilized in accordance with embodiments of the present disclosure. As shown in FIG. 6, computer system 600 can include an external storage device 610, a bus 620, a main memory 630, a read only memory 640, a mass storage device 650, communication port 660, and a processor 670. A person skilled in the art will appreciate that the computer system may include more than one processor and communication ports. Examples of processor 670 include, but are not limited to, an Intel® Itanium® or Itanium 2 processor(s), or AMD® Opteron® or Athlon MP® processor(s), Motorola® lines of processors, FortiSOC™ system on chip processors or other future processors. Processor 670 may include various modules associated with embodiments of the present invention. Communication port 660 can be any of an RS-232 port for use with a modem based dialup connection, a 10/100 Ethernet port, a Gigabit or 10 Gigabit port using copper or fiber, a serial port, a parallel port, or other existing or future ports. Communication port 660 may be chosen depending on a network, such a Local Area Network (LAN), Wide Area Network (WAN), or any network to which computer system connects. Memory 630 can be Random Access Memory (RAM), or any other dynamic storage device commonly known in the art. Read-only memory 640 can be any static storage device(s) e.g., but not limited to, a Programmable Read Only Memory (PROM) chips for storing static information e.g., start-up or BIOS instructions for processor 670. Mass storage 650 may be any current or future mass storage solution, which can be used to store information and/or instructions. Exemplary mass storage solutions include, but are not limited to, Parallel Advanced Technology Attachment (PATA) or Serial Advanced Technology Attachment (SATA) hard disk drives or solid-state drives (internal or external, e.g., having Universal Serial Bus (USB) and/or Firewire interfaces), e.g. those available from Seagate (e.g., the Seagate Barracuda 862 family) or Hitachi (e.g., the Hitachi Deskstar 8K600), one or more optical discs, Redundant Array of Independent Disks (RAID) storage, e.g. an array of disks (e.g., SATA arrays), available from various vendors including Dot Hill Systems Corp., LaCie, Nexsan Technologies, Inc. and Enhance Technology, Inc.
[0074] Bus 620 communicatively couples processor(s) 660 with the other memory, storage and communication blocks. Bus 620 can be, e.g. a Peripheral Component Interconnect (PCI) / PCI Extended (PCI-X) bus, Small Computer System Interface (SCSI), USB or the like, for connecting expansion cards, drives and other subsystems as well as other buses, such a front side bus (FSB), which connects processor 660 to software system.
[0075] Optionally, operator and administrative interfaces, e.g. a display, keyboard, and a cursor control device, may also be coupled to bus 620 to support direct operator interaction with a computer system. Other operator and administrative interfaces can be provided through network connections connected through communication port 660. The external storage device 610 can be any kind of external hard-drives, floppy drives, IOMEGA® Zip Drives, Compact Disc - Read Only Memory (CD-ROM), Compact Disc-Re-Writable (CD-RW), Digital Video Disk-Read Only Memory (DVD-ROM). Components described above are meant only to exemplify various possibilities. In no way should the aforementioned exemplary computer system limit the scope of the present disclosure.
[0076] Thus, the present disclosure provides a unique and inventive solution by providing a distributed edge and cloud system to assist with microlending services to communities, with machine learning catered to that specific community. A combination of technologies including microservices-based architecture and blockchain technology coupled with machine learning is utilized to provide microfinancing services to help sustain businesses in a local community, and to enable the community to grow into a thriving economy. To minimize the widespread expressed risk, the prediction of whether a loan will default or not is based on the various decision enabling parameters and on any available information about the borrowers’ past transaction as well as aggregate metrics related to the community that the borrower resides in.
[0077] While considerable emphasis has been placed herein on the preferred embodiments, it will be appreciated that many embodiments can be made and that many changes can be made in the preferred embodiments without departing from the principles of the invention. These and other changes in the preferred embodiments of the invention will be apparent to those skilled in the art from the disclosure herein, whereby it is to be distinctly understood that the foregoing descriptive matter to be implemented merely as illustrative of the invention and not as limitation.

ADVANTAGES OF THE PRESENT DISCLOSURE
[0078] The present disclosure provides for a system and a method for leveraging a combination of emerging technologies such as microservices, blockchain technology, and machine learning at an edge community to explore a microfinancing services application at the edge community.
[0079] The present disclosure provides for a system and a method for developing an edge model that could be hosted in the cloud or at an edge-server (if available) to cater to the needs of an edge community.
[0080] The present disclosure provides for a system and a method for including many edge communities and provide divergence in microlending services.
[0081] The present disclosure provides for a system and a method for utilizing edge learning with accurate predictions based on the nature of an edge community,
[0082] The present disclosure provides for a system and a method for leveraging other sustainable edge applications as well.

,CLAIMS:1. A system for assisting and managing microlending service, the system comprising:
one or more processors 202 coupled with a memory 204, wherein said memory 204 stores instructions which when executed by the one or more processors 202 causes said system to:
receive, a set of data packets, wherein the set of data packets pertain to information associated to a user involved in one of lending or borrowing funds, the set of data packets are received from a plurality of databases communicatively coupled to one or more servers;
extract, by a feature extraction engine 214, a first set of features from the set of data packets received, said first set of features pertaining to a plurality of predefined parameters associated with a loan status of the user;
determine, by an ML engine 216, a microlending model based on a predefined set of instructions and the first set of features extracted;
train, by the ML engine 216, the microlending model based on the predefined set of instructions, the first set of features extracted; and
based on the trained microlending model, predict, by the ML engine, a probability for the user to be eligible to borrow or lend a predicted amount of funds.
2. The system as claimed in claim 1, wherein the one or more servers 112 comprise one or more edge servers 112-1, one or more cloud servers 112-2 or a combination thereof.
3. The system as claimed in claim 1, wherein the plurality of predefined parameters comprise a default value, a charged off value, a fully paid value, a current value, an issued value, a grace period value, a first late value, a second late value or a combination thereof.
4. The system as claimed in claim 1, wherein the set of data packets is subjected to feature engineering by the feature extraction engine 214 to extract the first set of features having a high positive correlation with a target variable, wherein the target variable is associated with the probability for the user to be eligible to borrow or lend the predicted amount of funds.
5. The system as claimed in claim 4, wherein one or more target variables are converted into numeric values by categorizing and assigning weights to the target values defined based on the loan status.
6. The system as claimed in claim 1, wherein the first set of features is pre-processed to obtain a second set of features to compute a feature importance value for each of the second set of features.
7. The system as claimed in claim 1, wherein a cross validation module is applied on the first set of features in order to split and validate the set of data packets to obtain an adequate set of classes associated with the first set of features.
8. The system as claimed in claim 1, wherein a multiservice platform is operatively coupled to the processor to assist and manage microlending of a plurality of users across a plurality of distributed computing user devices associated with a blockchain.
9. The system as claimed in claim 8, wherein the multiservice platform 226 further comprises a lender interaction module, a borrower interaction module, one or more prediction models, Inter Planetary File System (IPFS) front end module, a blockchain front end module, the edge server and the cloud server.
10. The system as claimed in claim 8, wherein the multiservice platform is operatively coupled to the plurality of databases.
11. A method for assisting and managing microlending service, the method comprising:
receiving, by a processor, a set of data packets, wherein the set of data packets pertain to information associated to a user involved in one of lending or borrowing funds, the set of data packets are received from a plurality of databases communicatively coupled to one or more servers;
extracting, by a feature extraction engine 214, a first set of features from the set of data packets received, said first set of features pertaining to a plurality of predefined parameters associated with a loan status of the user;
determining, by an ML engine 216, a microlending model based on a predefined set of instructions and the first set of features extracted;
training, by the ML engine 216, the microlending model based on the predefined set of instructions, the first set of features extracted; and
based on the trained microlending model, predicting, by the ML engine 216, a probability for the user to be eligible to borrow or lend a predicted amount of funds.
12. The method as claimed in claim 11, wherein the one or more servers comprise one or more edge servers 112-1, one or more cloud servers 112-2 or a combination thereof.
13. The method as claimed in claim 11, wherein the plurality of predefined parameters comprise a default value, a charged off value, a fully paid value, a current value, an issued value, a grace period value, a first late value, a second late value or a combination thereof.
14. The method as claimed in claim 11, wherein the method further comprises:
feature engineering, by the feature extraction engine, the set of data packets data to feature engineering to extract the first set of features having a high positive correlation with a target variable, wherein the target variable is associated with the probability for the user to be eligible to borrow or lend the predicted amount of funds.
15. The method as claimed in claim 14, wherein the method further comprises:
converting one or more target variables into numeric values by categorizing and assigning weights to the target values defined based on the loan status.
16. The method as claimed in claim 11, wherein the method further comprises:
pre-processing the first set of features to obtain a second set of features to compute a feature importance value for each of the second set of features.
17. The method as claimed in claim 11, wherein the method further comprises:
applying across validation module on the first set of features in order to split and validate the set of data packets to obtain an adequate set of classes associated with the first set of features.
18. The method as claimed in claim 11, wherein a multiservice platform is operatively coupled to the processor to assist and manage microlending of a plurality of users across a plurality of distributed computing user devices associated with a blockchain.
19. The method as claimed in claim 18, wherein the multiservice platform 206 further comprises a lender interaction module, a borrower interaction module, one or more prediction models, Inter Planetary File Method (IPFS) front end module, a blockchain front end module, the edge server and the cloud server.
20. The method as claimed in claim 18, wherein the multiservice platform is operatively coupled to the plurality of databases.