A real-time data intelligent detection method, device and computer equipment

By generating Pattern objects using the Flink-CEP engine and Janino compiler, the complexity of the architecture and the static nature of rules in traditional risk control solutions are solved, enabling dynamic rule modification and multi-rule execution, meeting business needs, reducing costs and improving flexibility.

CN115220698BActive Publication Date: 2026-06-30SIJIBAO
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SIJIBAO
Filing Date
2022-03-18
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Traditional risk control solutions have complex architectures and high development costs, and data detection rules cannot be dynamically modified, making them unable to meet the needs of certain business scenarios.

Method used

It uses the Flink-CEP engine and the Janino compiler to generate Pattern objects to achieve condition matching and output. It supports multiple data source accesses and obtains the results of satisfied and untriggered events through the PatternAPI function, and dynamically modifies the rule conditions.

Benefits of technology

It simplifies the architecture logic, reduces development and maintenance costs, supports multiple rules running simultaneously, dynamically modifies rules, meets flexible business needs, and retains data output that does not meet the rules.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115220698B_ABST
    Figure CN115220698B_ABST
Patent Text Reader

Abstract

This invention provides a real-time data intelligent detection method, apparatus, and computer equipment. The method includes: receiving real-time streaming data from a Kafka messaging system; using the Java compiler Janino to transform rule conditions and generate a Pattern object; using PatternAPI functions to retrieve data from the input real-time streaming data that meets the conditions from the Pattern object, and obtaining the triggered event results through the select method; using PatternAPI functions to retrieve data from the input real-time streaming data that meets the conditions from the Pattern object, and obtaining the non-triggered event results through the select notMatches method. The advantages of this invention are: simple architecture logic, low development and maintenance costs; and meeting the needs of flexible rule changes in business operations.
Need to check novelty before this filing date? Find Prior Art