Search

CN-122022992-A - Abnormality processing method and device, storage medium, electronic equipment and product

CN122022992ACN 122022992 ACN122022992 ACN 122022992ACN-122022992-A

Abstract

The specification provides an exception handling method, an exception handling device, a storage medium, electronic equipment and a product, which are applied to a credit investigation system corresponding to a credit investigation organization. The method comprises the steps of inputting data to be detected into a preset anomaly detection model to obtain anomaly detection results of multiple detection dimensions output by the anomaly detection model, wherein the multiple detection dimensions are selected from fields contained in the data to be detected, performing logic contradiction detection on the data to be detected according to a time data sequence corresponding to the data to be detected, performing anomaly user behavior detection on the data to be detected based on user behavior characteristics of users, performing anomaly user behavior detection on the data to be detected based on group attribute characteristics of user groups to which the users belong, determining target weights corresponding to all detection dimensions, performing comprehensive calculation on the anomaly detection results of all detection dimensions according to the target weights to obtain comprehensive detection results for the data to be detected, and performing anomaly processing according to the comprehensive detection results.

Inventors

  • HU QUN

Assignees

  • 钱塘征信有限公司

Dates

Publication Date
20260512
Application Date
20260409

Claims (12)

  1. 1. An exception handling method, comprising: acquiring data to be detected provided by a data provider; Inputting the data to be detected into a preset abnormality detection model to obtain abnormality detection results of a plurality of detection dimensions output by the abnormality detection model, wherein the plurality of detection dimensions are selected from the group consisting of performing logic contradiction detection on fields contained in the data to be detected, performing abnormal value detection on the data to be detected according to a time data sequence corresponding to the data to be detected, performing abnormal user behavior detection on the data to be detected based on user behavior characteristics of a user, and performing abnormal user behavior detection on the data to be detected based on group attribute characteristics of a user group to which the user belongs; Determining target weights corresponding to all detection dimensions, comprehensively calculating abnormal detection results of all detection dimensions according to the target weights to obtain comprehensive detection results aiming at the data to be detected, and carrying out abnormal processing according to the comprehensive detection results.
  2. 2. The method of claim 1, wherein the abnormal user behavior detection is performed on the data to be detected based on user behavior characteristics of a user, and specifically comprises: determining behavior data of a user from the data to be detected, and determining user behavior characteristics of the user in a personal information field corresponding to the user; if the deviation between the current behavior characteristic reflected by the behavior data and the user behavior characteristic is larger than a preset deviation, determining that abnormal user behaviors exist in the data to be detected.
  3. 3. The method of claim 1, wherein the abnormal user behavior detection is performed on the data to be detected based on group attribute characteristics of the user group to which the user belongs, and specifically comprises: determining behavior data of a user from the data to be detected, and determining group attribute characteristics corresponding to a user group to which the user belongs; If the behavior data is determined to be not in accordance with the normal behavior characteristics of the user group to which the behavior data belongs according to the group attribute characteristics, determining that abnormal user behaviors exist in the data to be detected.
  4. 4. The method of claim 1, determining a target weight corresponding to each detection dimension, specifically comprising: And updating the basic weight preset for each detection dimension according to the abnormality degree score contained in the abnormality detection result of the detection dimension to obtain the target weight corresponding to the detection dimension.
  5. 5. The method of claim 4, wherein updating the basic weight set in advance for the detection dimension to obtain the target weight corresponding to the detection dimension specifically comprises: if the abnormality degree score of the data to be detected under the detection dimension is larger than a first preset score, the basic weight corresponding to the detection dimension is increased, and the target weight is obtained; If the abnormality degree score of the data to be detected in the detection dimension is smaller than a second preset score, reducing the basic weight corresponding to the detection dimension to obtain the target weight; if the abnormality degree score of the data to be detected in the detection dimension is not greater than the first preset score and is not less than the second preset score, determining the basic weight as the target weight; The first preset score is greater than the second preset score, and the target weight is positively correlated with the abnormality score when the abnormality score is greater than the first preset score or less than the second preset score.
  6. 6. The method of claim 1, the method further comprising: for each detection dimension, if the deviation between the abnormal detection result and the comprehensive detection result in the detection dimension is larger than a preset deviation, determining a loss value corresponding to the detection dimension according to the deviation between the abnormal detection result and the comprehensive detection result in the detection dimension; And taking the minimized loss value as an optimization target, and adjusting model parameters related to the detection dimension in the model to be detected.
  7. 7. The method of claim 1, wherein the exception handling is performed according to the comprehensive detection result, and specifically comprises: Determining an abnormal processing strategy matched with the comprehensive detection result, and determining reference alarm information matched with the comprehensive detection result in a preset alarm information base; Constructing a prompt word according to the reference alarm information, the abnormality processing strategy and the comprehensive detection result, and inputting the prompt word into a preset information generation model to prompt the information generation model, wherein the reference alarm information is taken as a reference, and the abnormality alarm information aiming at the comprehensive detection result is generated based on the abnormality processing strategy; And carrying out exception handling according to the exception alarm information.
  8. 8. The method of claim 1, further comprising, prior to exception handling based on the integrated test result: constructing an abnormal data tracing map according to participation objects involved in the data transfer process of the historical data; performing exception handling according to the comprehensive detection result, specifically including: If the data to be detected is abnormal based on the comprehensive detection result, determining a participation object associated with a data source of the data to be detected in the abnormal data traceability map, and performing abnormal processing on the participation object.
  9. 9. An exception handling apparatus, comprising: The acquisition module is used for acquiring the data to be detected provided by the data provider; The detection module is used for inputting the data to be detected into a preset abnormality detection model to obtain abnormality detection results of a plurality of detection dimensions output by the abnormality detection model, wherein the plurality of detection dimensions are selected from the group consisting of performing logic contradiction detection on fields contained in the data to be detected, performing abnormal value detection on the data to be detected according to a time data sequence corresponding to the data to be detected, performing abnormal user behavior detection on the data to be detected based on user behavior characteristics of a user, and performing abnormal user behavior detection on the data to be detected based on group attribute characteristics of a user group to which the user belongs; the processing module is used for determining target weights corresponding to all detection dimensions, comprehensively calculating abnormal detection results of all detection dimensions according to the target weights, obtaining comprehensive detection results aiming at the data to be detected, and carrying out abnormal processing according to the comprehensive detection results.
  10. 10. An electronic device comprising a processor, a memory for storing processor-executable instructions, wherein the processor is configured to implement the steps of the method of any one of claims 1-8 by executing the executable instructions.
  11. 11. A computer readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the steps of the method of any of claims 1-8.
  12. 12. A computer program product comprising computer programs/instructions which, when executed by a processor, implement the steps of the method of any of claims 1-8.

Description

Abnormality processing method and device, storage medium, electronic equipment and product Technical Field One or more embodiments of the present disclosure relate to the field of computer technology, and in particular, to an exception handling method, an exception handling apparatus, a storage medium, an electronic device, and a product. Background Along with the rapid development of credit services, the requirements of credit assessment on the credibility of data are remarkably improved, and accurate and reliable credit data become an important foundation for developing credit services and preventing credit risks for financial institutions. The source of credit investigation data is generally wide, and the data types are complex and various, so that abnormal detection of the data provided by the data source organization becomes a core link of the credit investigation organization for guaranteeing the reliability of credit investigation. However, due to the complexity of the credit investigation data, the form of generating the anomaly also presents concealment and diversity, so that the existing anomaly detection method is difficult to effectively cover anomalies in different forms, even causes missed detection and misjudgment of the anomaly, and the accuracy of anomaly detection is lower, so that the severe requirements of credit investigation business on data quality are difficult to be met. Disclosure of Invention In view of this, one or more embodiments of the present disclosure provide the following technical solutions: According to a first aspect of one or more embodiments of the present disclosure, an exception handling method is provided, which is applied to a credit investigation system corresponding to a credit investigation institution, and includes: acquiring data to be detected provided by a data provider of a credit investigation service; Inputting the data to be detected into a preset abnormality detection model to obtain abnormality detection results of a plurality of detection dimensions output by the abnormality detection model, wherein the plurality of detection dimensions are selected from the group consisting of performing logic contradiction detection on fields contained in the data to be detected, performing abnormal value detection on the data to be detected according to a time data sequence corresponding to the data to be detected, performing abnormal user behavior detection on the data to be detected based on user behavior characteristics of a user, and performing abnormal user behavior detection on the data to be detected based on group attribute characteristics of a user group to which the user belongs; Determining target weights corresponding to all detection dimensions, comprehensively calculating abnormal detection results of all detection dimensions according to the target weights to obtain comprehensive detection results aiming at the data to be detected, and carrying out abnormal processing according to the comprehensive detection results. According to a second aspect of one or more embodiments of the present specification, there is provided an exception handling apparatus, comprising: The acquisition module is used for acquiring the data to be detected provided by the data provider; The detection module is used for inputting the data to be detected into a preset abnormality detection model to obtain abnormality detection results of a plurality of detection dimensions output by the abnormality detection model, wherein the plurality of detection dimensions are selected from the group consisting of performing logic contradiction detection on fields contained in the data to be detected, performing abnormal value detection on the data to be detected according to a time data sequence corresponding to the data to be detected, performing abnormal user behavior detection on the data to be detected based on user behavior characteristics of a user, and performing abnormal user behavior detection on the data to be detected based on group attribute characteristics of a user group to which the user belongs; the processing module is used for determining target weights corresponding to all detection dimensions, comprehensively calculating abnormal detection results of all detection dimensions according to the target weights, obtaining comprehensive detection results aiming at the data to be detected, and carrying out abnormal processing according to the comprehensive detection results. According to a third aspect of one or more embodiments of the present specification, an electronic device is presented, comprising a processor, a memory for storing processor executable instructions, wherein the processor implements the steps of the above method by executing the executable instructions. According to a fourth aspect of one or more embodiments of the present description, a computer-readable storage medium is presented, having stored thereon computer instructions which, when executed by a pr