TWM661310U - Automated Review System - Google Patents

Abstract

This disclosure is about an automated review system that operates in a computer host. The system includes a communication module, an automated process module, and a management module. According to an implementation method of this disclosure, the automated review system interfaces with the client and the server through the communication module, and communicates with the automated process module and the management module, and the automated module and the management module are also communicated with each other.

Description

Automated review system

This disclosure is about a system for reviewing consumer financial services or products. Specifically, this disclosure is a computer system for automating loan review and generating loan contracts.

Consumer finance refers to the financial services provided by financial institutions to natural person customers, including deposits, loans, remittances, credit cards, financial product sales and other services. The characteristics of consumer finance services are short-term loans, high interest rates and more relaxed credit standards. In the case of urgent need for funds, consumer finance services have become the first choice for many people. However, in order to ensure that the borrower has sufficient repayment ability, financial institutions or banks must conduct relevant risk assessments through a complicated credit review process. Such a review process is very time-consuming, and factors such as cumbersome document requirements, lack of automation, unreliable data sources and security risks may lead to inaccurate assessments of the borrower's credit risk and financial status.

Although many software or systems have been developed to automate specific processes, data science and artificial intelligence technologies can be used to analyze large amounts of personal and financial data, and algorithms can be used to assist in decision-making to improve tax review efficiency. However, there are still many challenges in automated tax review systems, for example: excessive reliance on historical data and algorithmic models, ignoring unstructured information and special circumstances; when algorithmic models are trained based on biased data, it may lead to unfair treatment of certain groups.

In view of this, this disclosure provides an automated credit review system that can automatically assess the credit risk of borrowers and generate loan contracts based on the parameters, paths and judgment logic set by the user. At the same time, it can be used by financial institutions or banks to monitor the execution of the credit review process and make adjustments based on special conditions and circumstances, thereby improving the efficiency of credit review and the accuracy of its results.

The novel content is intended to provide a simplified summary of the present disclosure so that readers can have a basic understanding of the present disclosure. This novel content is not a complete overview of the present disclosure, and it is not intended to point out the important/key elements of the present disclosure or to define the scope of the present disclosure.

The present disclosure is intended to provide an automated review system that operates in a computer host and is used to generate a loan contract. According to an implementation of the present disclosure, the automated review system includes a communication module, an automated process module, and a management module, wherein the communication module is respectively interfaced with a client and a server to receive an identity image, a real-time facial image, a financial certificate, and a loan application from the client; the automated process module is communicatively connected to the communication module to automatically perform a review process based on the loan application; and the management module is respectively communicatively connected to the communication module and the automated process module to allow the server to manage the review process. In a specific implementation of the present disclosure, the review process includes: (a) reviewing the client according to the internal data review rules and generating a first review result; (b) verifying the identity of the client according to the first review result and generating a verification result, wherein the client is verified using an identity certificate image and a real-time facial image; (c) confirming the employee identity of the client according to the verification result and assigning a first weight according to the employee identity; (d) accessing external credit rating data related to the client to generate a second review result (e) calculating the loan conditions based on the first weight and the second weight; (f) evaluating the loan conditions of procedure (e) with the financial evidence of the client to generate the repayment capacity review result; (g) generating the anti-money laundering risk assessment result of the client based on the anti-money laundering list; and (h) generating a loan contract based on the loan conditions of procedure (e), the repayment capacity review result of procedure (f) and the anti-money laundering risk assessment result of procedure (g), and transmitting the loan contract to the client.

According to certain implementations of this disclosure, the internal data verification rules include whether the age is less than or equal to 60 years old, whether the person is a rejected customer, whether the person is an interested party, and whether the person has applied for a loan repeatedly. According to a specific implementation of this disclosure, in procedure (a), the automated process module accesses the record data related to the client in the server, and verifies the record data according to the internal data verification rules to generate a first verification result.

In one embodiment of the disclosure, the identity proof image includes personal information and a facial image; and in procedure (b), the automated process module compares the facial image with the real-time facial image to generate a verification result. In another embodiment of the disclosure, in procedure (b), the automated process module further uses optical character recognition technology to extract the personal information to establish the applicant profile.

According to the optional implementation method of this disclosure, the external credit assessment data includes telecommunications data, bad debt records, delayed repayment records, non-payment records, joint inquiry records, credit scores and work status.

According to the exemplary implementation of the present disclosure, the loan conditions include a loan amount and a loan interest rate. In an embodiment of the present disclosure, the automated process module calculates the loan interest rate based on the first weight and calculates the loan amount based on the second weight.

In the first embodiment of the present disclosure, in procedure (g), the client is compared with the anti-money laundering list to generate an anti-money laundering risk assessment result.

Optionally, the management module is also used for the server to modify the loan contract.

According to an implementation method of the disclosure, when the first check result of procedure (a), the verification result of procedure (b), the second check result of procedure (d), the repayment capacity review result of procedure (f), or the anti-money laundering risk assessment result of procedure (g) is not passed, the automated process module stops executing the review process.

According to another alternative implementation of the present disclosure, the management module is further used for the server to adjust the first check result of process (a), the verification result of process (b), and the second check result of process (d).

After reading the implementation method below, a person with ordinary knowledge in the technical field to which this new model belongs can easily understand the basic spirit and other new purposes of this new model, as well as the technical means and implementation methods adopted by this new model.

100: Automated review system

110: Communication module

120:Automated process module

130: Management module

S210-S280: Procedure

S: Server side

U: Client

In order to make the above and other purposes, features, advantages and implementation methods of the present invention more clearly understood, the attached drawings are described as follows.

Figure 1 is a schematic diagram of an automated review system 100 according to an implementation method of the present disclosure; and Figure 2 is a flow chart of the review process according to a specific implementation method of the present disclosure.

According to the usual practice, the various features and components in the figure are not drawn to scale. The drawing method is to present the specific features and components related to the new invention in the best way. In addition, the same or similar component symbols are used to refer to similar components/parts between different figures.

In order to make the description of the disclosed content more detailed and complete, the following provides an illustrative description of the implementation and specific embodiments of the new invention; however, this is not the only form of implementing or using the specific embodiments of the new invention. The implementation method covers the features of multiple specific embodiments and the method procedures and their sequences for constructing and operating these specific embodiments. However, other specific embodiments can also be used to achieve the same or equal functions and procedure sequences.

I Definition

For convenience, specific terms used in this specification, embodiments and the attached patent application are collected here. Unless otherwise defined in this specification, the meaning of scientific and technical terms used here is the same as that understood and used by ordinary knowledgeable people in the technical field to which this novel invention belongs. In addition, where there is no conflict with the context, singular terms used in this specification include plural forms of the terms, and plural terms used also include singular forms of the terms. Specifically, in this specification and the patent application, the singular form "a" (a and an) includes plural reference values, unless otherwise indicated by the context. In addition, in this specification and the patent application, the expressions "at least one" and "one or more" have the same meaning, both of which represent one, two, three or more.

In this disclosure, the term "interfacing" refers to the process of establishing a connection or integration between different devices, systems, applications or platforms. It allows two interfaced systems, applications, platforms or combinations thereof to transfer data, information or instructions to each other. "Interfacing" is generally based on standardized protocols, application programming interfaces (APIs), plug-ins, middleware, etc. to achieve data sharing, automated processes, cross-platform operations, etc.

In the present disclosure, a "server" or "client" includes any computer device capable of communicating with at least one server, wherein the communication connection is not limited to a wired or wireless network connection. According to an implementation method of the present invention, the "server" or "client" includes at least one graphical display device and a graphical user interface, allowing users to view messages and interact through applications, tools, services or software of the graphical user interface. The "server" or "client" is the location where the user interacts with the system through a computer device or terminal. In the present disclosure, the "server" is preferably an institution that provides consumer financial services, and the "client" is preferably an individual or institution that wishes to apply for consumer financial services. Furthermore, those with general knowledge in this technical field should understand that the number of the server or client can be one or more. In one embodiment, the number of the server and client is multiple.

In this disclosure, "financial proof" refers to documents or information used to prove the income, assets, debts and financial health of an individual or organization. Generally speaking, "financial proof" includes various income proofs, asset proofs, debt proofs, financial reports, tax documents or other documents.

In this disclosure, "identity image" refers to an image of a document issued by a government agency or a trusted organization that contains the user's personal information and facial photo and can be used to prove the user's identity and rights.

In this disclosure, "Anti-Money Laundering List" refers to a list compiled by government agencies, regulatory agencies or international organizations (such as the United Nations, the Financial Action Task Force (FATF), etc.), which lists individuals, entities or countries suspected of engaging in money laundering, financing terrorism or other illegal financial activities; or a list compiled by each enterprise based on business records. These lists are usually used by financial institutions, enterprises and regulatory agencies as a reference for assessing customer risks and performing regulatory responsibilities. According to an embodiment of this disclosure, the "Anti-Money Laundering List" is compiled by the service end itself.

II. Specific implementation methods

In order to improve the defects of the previous technology, this disclosure provides an automated review system that can automatically analyze data and documents and make decisions based on the decision rules set by the user, and set up a background management module for users to make manual decisions based on special circumstances, so that the evaluation results are more accurate.

FIG. 1 is a schematic diagram of an automated review system 100 according to a specific implementation method of the present disclosure. The automated review system 100 is executed on a computer device or a computer host. Specifically, it can be a server built into a computer host, or it can include a server to execute the interaction between modules. As shown in the figure, the automated review system 100 includes a communication module 110, an automated process module 120, and a management module 130, wherein the communication module 110 is respectively connected to the client U and the server S to receive the loan application and the documents required for the loan (i.e., identity proof image, real-time facial image, and financial proof) from the client U, and the operation instructions from the server S; the automated process module 120 is further connected to the communication module 110 to automatically execute the review process according to the loan application and related documents of the client U, and/or the operation instructions of the server S; and the management module 130 is connected to the communication module 110 and the automated process module 120 to manage the review process.

First, the client U interacts with the new automated review system 100 through any computer device or terminal (e.g., personal computer, smart phone, tablet computer, etc.) that can execute the system. In an embodiment of the present disclosure, the automated review system 100 is executed by a computer host in a communication network, so the aforementioned interaction refers to the establishment of a connection between the client U device and the automated review system 100. Specifically, the client U interfaces with the communication module 110 of the automated review system 100 to establish a connection. The automated review system 100 generates a user operation screen on the personal device of the client U and provides operation instructions to guide the client U to transmit the loan application and related certification documents required for review to the automated review system 100.

According to the regulations of different lending institutions and the type of loan applied for, different supporting documents are required for the lending institution (i.e., the service end S) to evaluate the applicant (i.e., the user end U). According to the optional implementation of the content of this disclosure, the required relevant documents include, but are not limited to, financial certificates (e.g., payroll, salary receipts, profit reports, tax returns, bank statements, investment certificates, real estate certificates, employer letters, employment certificates, company registration certificates), identity certificates (e.g., identity cards, driver's licenses, passports), marriage certificates, residence certificates, guarantee certificates or combinations thereof. The format of submitting the aforementioned certification documents depends on the application method, which may be to provide images by taking photos or scanning documents, or to submit paper files of the original or photocopy of the certification documents, or to directly upload electronic files (such as PDF files, word files, etc.) to the system. In the new automated tax review system 100, the user terminal U is required to provide identity proof images and real-time facial images by taking photos and/or scanning documents with a camera, and the financial proof is transmitted to the automated tax review system 100 in any of the above-mentioned ways according to the original file format (i.e., paper or electronic files).

The automated process module 120 automatically performs a credit review of the loan application (i.e., performs a review process) based on the loan application submitted by the client U and the decision-making procedures and rules defined by the server S. Accordingly, the automated process module 120 can be any computer module known in the art that can automatically perform data capture, document recognition, content comparison, and decision analysis functions according to pre-established business processes and their decision logic and rules, such as Camunda, Drools, IBM Operational Decision Manager, JBoss BRMS, OpenRules, Apache jBPM, FICO Decision Management Suite, Experian PowerCurve, Oracle Policy Automation, SAS Decision Manager, TIBCO BusinessEvents. The above are only a few common decision modules for illustration and are not intended to limit the scope of the new automated review system 100. According to an embodiment of the present disclosure, the automated process module 120 is Camunda.

Please refer to FIG. 1 and FIG. 2 at the same time, wherein FIG. 2 is a flowchart of the review process executed by the automated process module 120 of FIG. 1 according to a specific implementation method of the present disclosure. As shown in the figure, the review process includes the following procedures: Procedure S210: review the client according to the internal data review rules and generate a first review result; Procedure S220: verify the identity of the client according to the first review result and generate a verification result; Procedure S230: confirm the employee identity of the client according to the verification result and give a first weight according to the employee identity; Procedure S240: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S241: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S242: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S243: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S244: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S245: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S246: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S247: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S248: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S249: access external credit rating data related to the client to generate a second review result, and give a second weight according to the second review result; Procedure S250: access external credit rating data related to the client to generate a second review result, and give a second Procedure S250: Calculate loan conditions based on the first weight and the second weight; Procedure S260: Evaluate the loan conditions of procedure S250 with the financial proof of the client and generate a repayment capacity review result; Procedure S270: Generate the anti-money laundering risk assessment result of the client based on the anti-money laundering list; and Procedure S280: Generate a loan contract based on the loan conditions of procedure S250, the repayment capacity review result of procedure S260 and the anti-money laundering risk assessment result of procedure S270, and transmit the loan contract to the client.

In procedure S210, the automated process module 120 retrieves the record data of the specific user corresponding to the client U in the database of the server S according to the identity of the client U, and verifies the loan eligibility of the client U according to the internal data verification rules. The internal data verification rules are set by the server S according to the verification conditions within the loan institution and/or the regulations set by the government agency. According to a specific implementation method of the present disclosure, the internal verification rules include whether the age is less than or equal to 60 years old, whether it is a rejected customer, whether it is an interested party, and whether the loan application is repeated. For example, if the applicant has had business dealings with the service end (loans, cash advances, use of credit cards issued by the service end, etc.), but has not made repayments in accordance with the contract or has other violations, and is deemed by the service end to be a rejected customer, then the user is not eligible to obtain a loan from the service end. Accordingly, in this disclosure, the automated process module 120 will generate a first verification result after verifying the record data according to the internal data verification rules to indicate whether the user end U is eligible for a loan.

After obtaining the first verification result that the client U is eligible for a loan, in procedure S220, the automated process module 120 compares the user's facial image on the identity proof image with the real-time facial image to verify the identity of the client U and generate a verification result. Specifically, the automated process module 120 can identify and extract features from the face contained in the image to confirm whether the identity proof provided by the client U is the same person as the real applicant, and to confirm whether it is a living person. When the user's facial image on the identity proof image matches the real-time facial image, the identity verification is passed; when the user's facial image on the identity proof image does not match the real-time facial image, the identity verification is not passed. In an optional implementation, the automated process module 120 may also identify and extract features from faces in the video. According to an implementation of the present disclosure, the real-time facial image is a facial image generated by a user U using a personal device to take a real-time photo or selfie when submitting a loan application.

According to the optional implementation method of the present disclosure, the identity certificate image includes personal information and facial image. In process S220, the automated process module 120 is further used to extract personal information using optical character recognition technology to establish the applicant's profile for subsequent retrieval by the server S.

Next, in process S230, it is confirmed whether the client U has an employee identity according to the verification result of process S220, and the first weight is given accordingly. After confirming the identity of the client U through procedure S220, the automated process module 120 retrieves the employee list containing all the personnel information from the database of the server S, and compares the user name of the client U or any information that can represent its identity (e.g., ID number, passport number) with the employee list. If any of the information is the same as the personnel information in the employee list, the employee identity of the client U is determined to be positive, and the employee preferential interest rate is given accordingly; if the identity information of the client U is not the same as the personnel information in the employee list, the employee identity of the client U is negative, and the general loan interest rate is given; wherein, the employee preferential interest rate or the general loan interest rate is the aforementioned first weight.

In order to understand the overall credit status of the client U, in procedure S240, the automated process module 120 is used to access external credit rating data related to the client U (credit rating data relative to the record data held internally by the server S), check the credit status of the client U based on the external credit rating data and generate a second check result, and give a second weight based on the second check result. Specifically, the server S pre-sets the source of external data. When the identity verification of the client U is completed, the automated process module 120 will automatically visit a specific website or database based on the preset source to search for credit rating data related to the client U, and then analyze the credit rating data based on the algorithm, and calculate the second weight used to calculate the loan conditions.

According to the optional implementation method of the present disclosure, the external credit assessment data includes, but is not limited to, telecommunications data, credit history (bad debt record, delayed repayment record, non-payment record, joint record), financial account information (savings account, investment account information and balance status), employment status (employment status, occupation, work history, income level), marital status, collateral information (property, car, etc. ownership or appraisal report), legal records (debt litigation records, bankruptcy records, court judgment records), and credit scores. In an embodiment of the present disclosure, the external credit assessment data includes telecommunications data, bad debt records, delayed repayment records, non-payment records, joint inquiry records, credit scores and work status.

After completing the verification and evaluation procedures of procedures S210-S240, in procedure S250, the automated process module 120 calculates the loan conditions that meet the requirements of the user terminal U based on the first weight and the second weight. According to the implementation method of the first disclosure, the loan conditions include the loan amount and the loan interest rate. In a specific implementation method of the first disclosure, the automated process module calculates the loan interest rate based on the first weight and calculates the loan amount based on the second weight.

According to the optional implementation method of the present disclosure, the automated process module 120 calculates the customer's monthly payment rate (MPR) and the maximum loan amount that can be afforded, and takes the lower one as the loan condition.

Next, in procedure S260, the financial certificate provided by the user U is compared with the loan conditions to assess whether the user U is able to bear the aforementioned loan conditions, thereby generating a repayment capacity review result.

In procedure S270, the identity of the client U is further compared with the anti-money laundering list to assess the money laundering risk of the client U and generate an anti-money laundering risk assessment result.

Finally, in process S280, a loan contract is generated based on the loan conditions of process S250, the repayment ability review results of process S260, and the anti-money laundering risk assessment results of process S270, and the loan contract is transmitted to the user terminal U.

It is worth noting that when the first check result of procedure S210, the verification result of procedure S220, the second check result of procedure S240, the repayment ability review result of procedure S260 or the anti-money laundering risk assessment result of procedure S270 is not passed, the automated process module 120 will stop executing the review process and generate a rejection notification to be sent to the client U.

Please refer to Figure 1 again. The management module 130 is used by the server S to manage the aforementioned review process. Specifically, the management module 130 can be used to (1) control management permissions so that authorized server S users can enter and change system settings; (2) allow the server S to view client data; (3) view the loan data provided by the client U, including application status, loan qualification verification results (i.e., first verification results), identity verification results (i.e., verification results), verification results of external credit assessment data (i.e., second verification results), review process related information, service call records, documents and attachments; (4) modify loan contracts; (5) manage the status of the review process; and (6) generate daily and monthly reports.

In an implementation of the present disclosure, the state of the management review process (point (5)) specifically refers to when the review process stops due to the failure of the first check result of procedure S210, the verification result of procedure S220 and/or the second check result of procedure S240, the server S can perform manual review through the management module 130 so that the automated process module 120 can continue to execute the subsequent review process. According to an optional implementation of the present disclosure, the automated process module 120 can turn on the manual processing mode so that the server S can perform manual review in a specific process.

It should be noted that, according to the specific implementation of the disclosed content, the automated review system 100 can be built into a server of a computer host, or can also include a server to execute the interaction between the aforementioned modules to complete the review process.

In summary, compared with the current automated decision-making system, the automated review system disclosed herein allows users to respond in real time according to the actual situation and conduct manual review to eliminate the bias caused by the biased data structure in the model construction process, which may harm the interests of the applicant. Through automated review, the rigorous review process including different stages can be completed more quickly and accurately, thereby saving the time cost required for the manual review process.

It should be understood that the above description of the implementation method is only given in the form of an embodiment, and those with ordinary knowledge in the art to which this field belongs can make various modifications. The above specification, embodiments and experimental results provide a complete description of the structure and use of the exemplary implementation method of the new model. Although various specific embodiments of the new model are disclosed in the above embodiments, they are not used to limit the new model. Those with ordinary knowledge in the art to which this model belongs can make various changes and modifications to it without deviating from the principle and spirit of the new model. Therefore, the scope of protection of the new model shall be based on the scope defined by the attached patent application.

100: Automated review system

110: Communication module

120:Automated process module

130: Management module

S: Server side

U: Client

Claims (12)

An automated review system is operated in a computer host and is used to generate a loan contract, comprising: a communication module, which is respectively connected to a client and a server, and is used to receive an identity proof image, a real-time face image, a financial certificate and a loan application from the client; an automated process module, which is connected to the communication module and is used to automatically execute a review process according to the loan application, wherein the The review process includes: (a) reviewing the client according to an internal data review rule and generating a first review result; (b) verifying the identity of the client according to the first review result and generating a verification result, wherein the identity proof image and the real-time facial image are used to verify the client; (c) confirming the employee identity of the client according to the verification result and granting a first authorization according to the employee identity. (d) accessing an external credit rating data related to the client to generate a second check result, and assigning a second weight according to the second check result; (e) calculating a loan condition according to the first weight and the second weight; (f) using the loan condition of the financial proof evaluation procedure (e) of the client to generate a repayment ability review result; (g) generating the user according to an anti-money laundering list; (h) generating the loan contract according to the loan conditions of procedure (e), the repayment capacity review result of procedure (f) and the anti-money laundering risk assessment result of procedure (g), and transmitting the loan contract to the client; and a management module, which is respectively connected to the communication module and the automated process module for the server to manage the review process. The automated review system as described in claim 1, wherein in procedure (a), the automated process module accesses a record data related to the client in the server, and checks the record data according to the internal data checking rule to generate the first checking result. The automated review system as described in claim 2, wherein the internal data review rules include whether the age is less than or equal to 60 years old, whether the person is a rejected customer, whether the person is an interested party, and whether the person has applied for a loan repeatedly. The automated verification system as described in claim 1, wherein the identity proof image includes personal information and a facial image; in procedure (b), the automated process module compares the facial image with the real-time facial image to generate the verification result. The automated review system as described in claim 4, wherein in procedure (b), the automated process module is further used to extract the personal information using optical character recognition technology to create an applicant profile. An automated credit review system as described in claim 1, wherein the external credit assessment data includes telecommunications data, non-performing loan records, delayed repayment records, non-payment records, joint inquiry records, credit scores, and work status. An automated review system as described in claim 1, wherein the loan conditions include a loan amount and a loan interest rate. An automated review system as described in claim 7, wherein the automated process module calculates the loan interest rate based on the first weight and calculates the loan amount based on the second weight. An automated vetting system as described in claim 1, wherein in process (g), the client is compared with the anti-money laundering list. The automated review system as described in claim 1, wherein the management module is further used for the server to modify the loan contract. The automated review system as described in claim 1, wherein when the first review result of procedure (a), the verification result of procedure (b), the second review result of procedure (d), the repayment capacity review result of procedure (f), or the anti-money laundering risk assessment result of procedure (g) is not passed, the automated process module stops executing the review process. The automated review system as described in claim 1, wherein the management module is further used for the server to adjust the first check result of process (a), the verification result of process (b), and the second check result of process (d).