An event processing method, apparatus, device and computer readable storage medium
By determining the event pattern and counting threshold in Apache Flink and generating target event matching logic, the problem of counting only continuous inflow events with a fixed time sequence is solved, and counting non-continuous inflow events is realized, improving applicability and compatibility with business scenarios.
Patent Information
- Application Number
- CN202210745281.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-27
- Publication Date
- 2026-07-17
- Estimated Expiration
- 2042-06-27
AI Technical Summary
Apache Flink can only count events that flow in continuously according to a fixed time sequence, which has poor applicability.
By defining event patterns and counting thresholds, target event matching logic is generated, enabling the counting of events that flow in continuously or discontinuously according to the target time sequence. Nondeterministic loosely continuous repeating count analysis is performed using the Repeat Until template and the Flink CEP dedicated library.
It enriches the counting semantics of complex event correlation analysis in Apache Flink, improves its applicability, enables the handling of non-continuous event inflows, and enhances compatibility and resilience in business scenarios.
Smart Images

Figure CN115145746B_ABST
Abstract
Citation Information
Patent Citations
Detection method for complex events in mass disordered data streams of Internet of Things Manufacturing
CN104408142A
Event processing method and device, equipment and machine readable storage medium
CN113296933A
Event processing method, device and equipment and computer readable storage medium
CN115168070A