CN-114529399-B - User data processing method, device, computer equipment and storage medium

CN114529399BCN 114529399 BCN114529399 BCN 114529399BCN-114529399-B

Abstract

The present application relates to the field of artificial intelligence technologies, and in particular, to a user data processing method, apparatus, computer device, and storage medium. The method comprises the steps of obtaining user data to be processed, wherein the user data to be processed comprises initial user data of different time points, conducting map construction on the initial user data of each time point to obtain a plurality of static knowledge maps, conducting feature extraction on the static knowledge maps to obtain static features corresponding to the static knowledge maps, wherein the static features are used for representing feature data fusing the node features in the static knowledge maps, connecting the static features corresponding to each time point in series to obtain dynamic features, and obtaining target feature data corresponding to the user data to be processed according to the dynamic features. By adopting the method, the user data to be processed can be accurately evaluated.

Inventors

XUE YUSHAN

Assignees

中国工商银行股份有限公司
中国工商银行股份有限公司

Dates

Publication Date: 20260421
Application Date: 20220218
Priority Date: 20220218

Claims (12)

1. A method of user data processing, the method comprising: Acquiring user data to be processed, wherein the user data to be processed comprises initial user data at different time points; extracting the initial user data of each time point to obtain triple data; Performing map construction based on the triplet data to obtain a plurality of static knowledge maps, wherein one time point corresponds to one static knowledge map; When the label is missing in the nodes in the static knowledge graph, reading the triplet data of the label missing nodes, and supplementing the label missing nodes according to the labels of the triplet data; Respectively extracting features of the plurality of static knowledge patterns to obtain static features corresponding to the plurality of static knowledge patterns, wherein the static features are used for representing and fusing feature data of each node feature in the static knowledge patterns; the static features corresponding to the time points are connected in series to obtain dynamic features; and obtaining target feature data corresponding to the user data to be processed according to the dynamic features.
2. The method of claim 1, wherein the feature extraction of the plurality of static knowledge-maps to obtain static features corresponding to the plurality of static knowledge-maps includes: Obtaining feature matrixes corresponding to the static knowledge maps according to the static knowledge maps; And carrying out feature fusion on each feature matrix to obtain static features corresponding to a plurality of static knowledge maps.
3. The method according to claim 1, wherein the obtaining target feature data corresponding to the user data to be processed according to the dynamic feature includes: Extracting the dynamic characteristics to obtain retention characteristics; Updating the retention feature to obtain an updated feature; And calculating according to the updated characteristics to obtain the target characteristic data.
4. The method according to claim 1, wherein the feature extraction of the plurality of static knowledge-maps to obtain static features corresponding to the plurality of static knowledge-maps is performed by a first model trained in advance; And obtaining target characteristic data corresponding to the user data to be processed according to the dynamic characteristics through a pre-trained second model.
5. The method of claim 4, wherein the training process of the first model and the second model comprises: Acquiring sample data, wherein the sample data carries labeling data and characteristic labels; inputting the sample data into a first model, and extracting the sample data through the first model to obtain sample characteristics; Calculating a first target loss function according to the sample characteristics and the characteristic labels, wherein the first target loss function is used for optimizing the first model until the first model is trained; Inputting the sample characteristics into a second model to predict the sample characteristics through the second model so as to obtain sample target characteristics of the sample characteristics; And calculating a second target loss function according to the sample target characteristics and the labeling data, wherein the second target loss function is used for optimizing the second model until the second model is trained.
6. A risk handling method, the method comprising: Acquiring user data to be processed corresponding to a user to be predicted; obtaining target feature data corresponding to the user data to be processed according to the method of any one of claims 1-5; and determining the risk level of the user to be predicted according to the target characteristic data.
7. The method of claim 6, wherein said determining the risk level of the user to be predicted from the target feature data comprises: acquiring a preset evaluation grade; and obtaining the risk grade of the user to be predicted according to the target characteristic data and the evaluation grade.
8. A risk assessment device, the device comprising: the data acquisition module is used for acquiring user data to be processed, wherein the user data to be processed comprises initial user data at different time points; The map construction module is used for carrying out map construction on the initial user data of each time point to obtain a plurality of static knowledge maps, and one time point corresponds to one static knowledge map; The feature extraction module is used for respectively carrying out feature extraction on the plurality of static knowledge patterns to obtain static features corresponding to the plurality of static knowledge patterns, wherein the static features are used for representing and fusing feature data of each node feature in the static knowledge patterns; the feature processing module is used for connecting static features corresponding to the time points in series to obtain dynamic features; The target feature calculation module is used for obtaining target feature data corresponding to the user data to be processed according to the dynamic features; The map construction module is specifically configured to: extracting the initial user data of each time point to obtain triple data; performing map construction based on the triplet data to obtain a plurality of static knowledge maps; And the map construction module is also used for reading the triplet data where the label missing node is located when the label missing exists in the nodes in the static knowledge map, and supplementing the label missing node according to the label of the triplet data.
9. A risk handling device, the device comprising: the to-be-processed user data acquisition module is used for acquiring to-be-processed user data corresponding to a user to be predicted; A risk prediction module, configured to obtain target feature data corresponding to the user data to be processed according to the apparatus of claim 8; And the risk level judging module is used for determining the risk level of the user to be predicted according to the target characteristic data.
10. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 7 when the computer program is executed.
11. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.
12. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.

Description

User data processing method, device, computer equipment and storage medium Technical Field The present application relates to the field of artificial intelligence, and in particular, to a user data processing method, apparatus, computer device, storage medium, and computer program product. Background Loans are credit activities that carry out monetary funds at a certain interest rate and a defined period, and are the most important items in commercial banking. In which the level of profits of a loan has a direct relationship with the price of the loan, in particular, the price of the loan is high, the profits are high, but the demand of the loan will be reduced accordingly. Conversely, the loan price is low, the profit is low, but the loan demand will increase. Accordingly, there is a need to formulate corresponding loan interest rates for different lenders. In the related art, for example, in the 40 th century of 20, some banks in the united states have started to try to study credit scoring methods for rapidly processing a large number of credit applications, in 1956, engineers BillFair and mathematicians EARLISAAC have invented a well-known FICO scoring method, which uses a logistic regression method as a technical core and is the most mature credit risk scoring model applied in the current industry, and in the 60 th to 80 th century, along with the progress of information technology and the rapid development of business, the credit scoring model is widely applied to credit cards, consumed credit, house mortgage loans and small enterprise loans. However, these methods have many limitations in the big data age of data explosion, resulting in inefficiency and inaccurate evaluation results. Disclosure of Invention In view of the foregoing, it is desirable to provide a user data processing method, apparatus, computer device, and storage medium capable of accurately evaluating user data to be processed. In a first aspect, the present application provides a user data processing method, the method comprising: acquiring user data to be processed, wherein the user data to be processed comprises initial user data at different time points; Carrying out map construction on the initial user data of each time point to obtain a plurality of static knowledge maps; respectively extracting features of the plurality of static knowledge patterns to obtain static features corresponding to the plurality of static knowledge patterns, wherein the static features are used for representing feature data of each node feature in the fused static knowledge patterns; The static features corresponding to the time points are connected in series to obtain dynamic features; and obtaining target feature data corresponding to the user data to be processed according to the dynamic features. In one embodiment, the performing the graph construction on the initial user data at each time point to obtain a plurality of static knowledge graphs includes: extracting initial user data of each time point to obtain triplet data; and carrying out map construction based on the triplet data to obtain a plurality of static knowledge maps. In one embodiment, after the performing the graph construction based on the triplet data to obtain a plurality of static knowledge graphs, the method includes: When the label is missing in the nodes in the static knowledge graph, the triplet data of the label missing nodes are read, and the label missing nodes are supplemented according to the labels of the triplet data. In one embodiment, the extracting features of the plurality of static knowledge patterns to obtain static features corresponding to the plurality of static knowledge patterns includes: Obtaining a feature matrix corresponding to the plurality of static knowledge patterns according to the plurality of static knowledge patterns; and carrying out feature fusion on each feature matrix to obtain static features corresponding to the plurality of static knowledge maps. In one embodiment, the obtaining the target feature data corresponding to the user data to be processed according to the dynamic feature includes: extracting the dynamic characteristics to obtain retention characteristics; updating the reserved characteristics to obtain updated characteristics; and calculating according to the updated characteristics to obtain target characteristic data. In one embodiment, the feature extraction of the plurality of static knowledge patterns to obtain the static features corresponding to the plurality of static knowledge patterns is implemented through a first model trained in advance; the target feature data corresponding to the user data to be processed is obtained according to the dynamic features through a pre-trained second model. In one embodiment, the training process of the first model and the second model includes: Acquiring sample data, wherein the sample data carries labeling data and characteristic labels; Inputting the sample data into a first model to