Abnormal call record processing method, device, equipment, medium and product

By constructing initial exception rules and using a multimodal rule engine to analyze call detail records (CDRs), and dynamically updating the exception rules, the problem that static rule bases cannot adapt to new exception types is solved, and more efficient exception CDR identification and processing is achieved.

CN122388618APending Publication Date: 2026-07-14CHINA MOBILE GRP GUANGDONG CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610349481.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-20
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

In existing technologies, static rule bases cannot adapt to the billing logic of different abnormal call detail records (CDRs), resulting in insufficient flexibility in abnormal CDR identification, especially when new abnormal types appear.

Method used

By acquiring the characteristics of initial abnormal call records, initial abnormal rules are constructed, and static, dynamic, and time-series rules are used to analyze the call records to be audited. If the matching fails, the second abnormal feature is extracted to update the rules, including the number of call records and the date, to achieve dynamic adaptation.

Benefits of technology

It improves the flexibility and accuracy of abnormal call detail record (CDR) identification, reduces the cost of manual rule maintenance, shortens the time for manual review, and covers minor anomalies that are overlooked in traditional methods.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122388618A_ABST
    Figure CN122388618A_ABST
Patent Text Reader

Abstract

The present disclosure relates to the technical field of communication, in particular to an abnormal call record processing method, device, equipment, medium and product, the method comprising: obtaining a plurality of initial abnormal call records; extracting a first abnormal feature in each of the initial abnormal call records, and constructing an abnormal rule based on the first abnormal feature; wherein the first abnormal feature comprises a call record type, an error code and at least one charging identifier field; analyzing a to-be-audited abnormal call record through at least one preset target rule to obtain an analysis result; in the case that the analysis result and the abnormal rule fail to match, extracting a second abnormal feature in the to-be-audited abnormal call record, and updating the abnormal rule based on the second abnormal feature; wherein the second abnormal feature comprises the first abnormal feature, a call record quantity and a call record date. Through the target rule, it is judged whether the to-be-audited abnormal call record is an unknown abnormal mode, and dynamic adaptation of the abnormal rule is realized.
Need to check novelty before this filing date? Find Prior Art