CN-122022989-A - Recommendation method and device for credit agency

CN122022989ACN 122022989 ACN122022989 ACN 122022989ACN-122022989-A

Abstract

The application relates to a recommendation method and device of credit institutions, the method comprises the steps of obtaining model scores of target users on each credit institution, wherein the model scores are used for indicating the probability of credit institutions for giving credit to the target users, obtaining credit correlation among a plurality of credit institutions, wherein the credit correlation is related to a first matrix and a second matrix, elements in the first matrix are model scores of different users on one of the plurality of credit institutions, elements in the second matrix are model scores of different users on the other credit institution, classifying the plurality of credit institutions according to the credit correlation, and selecting a target number of credit institutions from each category to form a target credit institution combination for recommendation to the target users. The application solves the problem that the credit assisting platform in the related technology lacks a mature mechanism recommendation system, and realizes the effect of improving the accuracy of recommending credit mechanisms for users and further improving the viscosity and the dependence of the users on the credit platform.

Inventors

CHEN AO
SUN RUI

Assignees

上海旭荣网络科技有限公司

Dates

Publication Date: 20260512
Application Date: 20230328

Claims (10)

1. A method of recommending credit institutions, comprising: obtaining a model score of a target user on each credit agency, wherein the model score is used for indicating the probability of credit agency credit giving credit to the target user; Obtaining credit correlation between a plurality of credit mechanisms, wherein the credit correlation is related to a first matrix and a second matrix, elements in the first matrix are model scores of different users in one of the plurality of credit mechanisms, and elements in the second matrix are model scores of different users in another of the plurality of credit mechanisms; Classifying the plurality of credit authorities according to the trust relevance; selecting a target number of credit agencies from within each category to compose a target credit agency combination recommendation to the target user.
2. The method of claim 1, wherein obtaining trust correlations between a plurality of credit agencies comprises: Calculating a credit correlation rho (a, B) between credit agency a and credit agency B according to the following formula: Wherein ,u=[score(A1), score(A2), score(A3),...,score(An)], v=[score(B1), score(B2), score(B3),...,score(Bn)], the score (An) is used to represent the model score of the nth user on the credit agency a and the score (Bn) is used to represent the model score of the nth user on the credit agency B.
3. The credit agency recommendation method of claim 1, wherein categorizing the plurality of credit agencies according to the credit correlation includes: grouping credit agencies with credit correlation exceeding a first value in the credit agencies in pairs into a first category; classifying the trust authorities with trust relativity lower than a second value with other trust authorities in the trust authorities into a second class; And for the rest credit giving mechanisms except the first category and the second category in the plurality of credit giving mechanisms, if the credit giving correlation between the rest credit giving mechanisms and the credit giving mechanisms exceeding the target number in the first category exceeds the first numerical value, classifying the rest credit giving mechanisms into the first category, otherwise classifying the rest credit giving mechanisms into a third category.
4. The method of claim 1 wherein selecting a target number of credit mechanisms within each category to make up a target credit mechanism combination recommendation to the target user comprises: Selecting the target number of credit mechanisms from each category to form a target credit mechanism combination according to the ordering of the credit probability of the credit mechanisms in each category; recommending the target credit agency combination to the target user.
5. The method of recommending credit mechanisms according to any one of claims 1 to 4, further comprising, after selecting a target number of credit mechanism component target credit mechanism combinations from within each category for recommending to the target user: calculating an overall credit probability p (a, B) for the target user over the target credit agency combination, wherein the target credit agency combination includes credit agency a and credit agency B, according to the following formula: where p 1 represents the probability of credit for the target user at credit agency A and p 2 represents the probability of credit for the target user at credit agency B.
6. A credit agency recommendation device, comprising: A first obtaining unit, configured to obtain a model score of a target user on each credit agency, where the model score is used to indicate a probability that a credit agency trusts the target user; A second obtaining unit, configured to obtain a credit correlation between a plurality of credit mechanisms, where the credit correlation is related to a first matrix and a second matrix, elements in the first matrix are model scores of different users in one of the plurality of credit mechanisms, and elements in the second matrix are model scores of different users in another of the plurality of credit mechanisms; A classification unit for classifying the plurality of credit authorities according to the credit correlation; A recommending unit, configured to select a target number of credit mechanisms from each category to compose a target credit mechanism combination for recommending to the target user.
7. The credit agency recommendation device of claim 6, wherein the second acquisition unit includes: a calculation module for calculating a credit correlation rho (a, B) between credit agency a and credit agency B according to the following formula: Wherein ,u=[score(A1), score(A2), score(A3),...,score(An)], v=[score(B1), score(B2), score(B3),...,score(Bn)], the score (An) is used to represent the model score of the nth user on the credit agency a and the score (Bn) is used to represent the model score of the nth user on the credit agency B.
8. The credit agency recommendation device of claim 6, wherein the classification unit includes: The first classification module is used for classifying credit institutions with the credit correlation exceeding a first value in the plurality of credit institutions into a first class; The second classification module is used for classifying the trust mechanisms with trust relativity lower than a second value with the rest trust mechanisms in the plurality of trust mechanisms into a second class; and the third classification module is used for classifying the remaining credit mechanisms except the first category and the second category in the plurality of credit mechanisms, classifying the remaining credit mechanisms into the first category if the credit correlation between the remaining credit mechanisms and the credit mechanisms exceeding the target number in the first category exceeds the first numerical value, or classifying the remaining credit mechanisms into the third category.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the recommendation method of a credit mechanism as claimed in any one of claims 1 to 5 when the computer program is executed.
10. A computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a credit agency recommendation method according to any one of claims 1 to 5.

Description

Recommendation method and device for credit agency The application relates to a filing application of a recommending method and a recommending device for a credit agency, which is applied for the filing application of 2023, 3, 28 and 202310311172. X. Technical Field The present invention relates to the field of financial credit technology, and in particular, to a credit agency recommendation method, apparatus, computer device, and computer readable storage medium. Background In the field of financial credit, especially in the field of internet micro-loans, loan applicants often butt-joint financial institutions through some application programs, software and a loan assisting platform, and the institutions or the loan assisting platform determine whether to credit the applicant and credit line and charge by checking the qualification of the applicant. In practical business applications, a considerable number of users apply for loans by downloading apps to a lending platform, which may be connected to a plurality of small lending institutions or sponsors to assist the users in obtaining the loans. By this mechanism, it is possible to interface with multiple financial institutions by the lending user only needing to download an App. The preferences of different financial institutions for users and risks are naturally different, so that the same user has different claims, deadlines and fees at different institutions, and the user has more options. In the whole flow, the lending platform is used as a flow party, certain auditing and wind control can be carried out on the qualification of the user, the unconditional user cannot be pushed to the institution, and a recommendation system is established on the lending platform for the eligible user, so that the user can obtain a loan through the lending platform. At present, no mature recommendation system exists for the recommended strategy of institutions on a lending assistance platform. In a practical level, the lending platform generally faces several bottlenecks or problems as follows: 1. how can what users meet the requirements of an organization? 2. How do the preferences of the institutions and the relevance of preferences between institutions quantify? 3. How do the same user compare the possibilities of trust in different institutions? 4. According to the probability of trust and the preference of the organization, a set of strategies is formulated so that the rate of trust (rate of quota) can be improved? At present, no effective solution is proposed for the problem that the loan-aid platform in the related art lacks a mature institution recommendation system. Disclosure of Invention The application aims at overcoming the defects in the prior art and provides a credit agency recommendation method, a credit agency recommendation device, a credit agency recommendation computer device and a credit agency recommendation computer-readable storage medium, so as to at least solve the problem that a credit assisting platform in the related art lacks a mature agency recommendation system. In order to achieve the above purpose, the technical scheme adopted by the application is as follows: in a first aspect, an embodiment of the present application provides a credit agency recommendation method, including: obtaining a model score of a target user on each credit agency, wherein the model score is used for indicating the probability of credit agency credit giving credit to the target user; Obtaining credit correlation between a plurality of credit mechanisms, wherein the credit correlation is related to a first matrix and a second matrix, elements in the first matrix are model scores of different users in one of the plurality of credit mechanisms, and elements in the second matrix are model scores of different users in another of the plurality of credit mechanisms; Classifying the plurality of credit authorities according to the trust relevance; selecting a target number of credit agencies from within each category to compose a target credit agency combination recommendation to the target user. In some of these embodiments, obtaining trust correlations between a plurality of credit authorities includes: Calculating a credit correlation rho (a, B) between credit agency a and credit agency B according to the following formula: Wherein ,u = [score(A1), score(A2), score(A3),...,score(An)], v = [score(B1), score(B2), score(B3),...,score(Bn)], the score (An) is used to represent the model score of the nth user on the credit agency a and the score (Bn) is used to represent the model score of the nth user on the credit agency B. In some of these embodiments, categorizing the plurality of credit authorities according to the trust relevance includes: grouping credit agencies with credit correlation exceeding a first value in the credit agencies in pairs into a first category; classifying the trust authorities with trust relativity lower than a second value with other trust aut