Complex Event Processing Rule Generation via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current complex event processing systems require manual rule creation by human operators or machine learning from single input events, and cannot automatically generate rules for streams of multiple events to infer output events.
Innovation Solution
A method using standard learning algorithms to create input event feature vectors and generate rules for complex event processing systems, enabling automated rule generation for both predefined output events and anomaly detection in input event streams.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual rule creation by human operators is used, then rule accuracy and system reliability are improved, but device complexity and operational burden increase
Solution Approach 1:
The system automatically generates CEP rules by analyzing event streams and identifying patterns, allowing the system to serve itself without requiring manual rule creation by human operators. The rule generation process is fully automated through machine learning algorithms that process event data and produce actionable rules.
Solution Approach 2:
Manual mechanical rule creation processes are replaced with automated machine learning algorithms that compute rules from event stream data. The system uses computational methods to analyze patterns in events and generate rules, substituting human operational burden with automated mechanical processing.
2Extent of automation
If machine learning from single input events is used, then automation extent is improved, but the ability to handle complex multi-event scenarios deteriorates
Solution Approach 1:
The event stream is segmented into discrete events, and the system analyzes sequences of these segmented events to identify patterns. This segmentation allows the machine learning algorithm to process complex multi-event scenarios by breaking them down into manageable individual events and their temporal relationships.
Solution Approach 2:
The system performs preliminary analysis of event streams to identify patterns and relationships before generating final rules. By pre-processing the event data and identifying temporal patterns in advance, the system prepares the foundation for accurate rule generation from multiple events.
3Productivity
If automated rule generation for multiple events is implemented, then productivity and system response time are improved, but measurement precision and rule accuracy may deteriorate
Solution Approach 1:
The system incorporates feedback mechanisms where generated rules are validated against actual event stream data to ensure accuracy. The machine learning algorithm continuously refines rule generation based on feedback from observed event patterns, maintaining high measurement precision while achieving automated productivity.
Solution Approach 2:
The system dynamically adjusts parameters such as time windows, event thresholds, and pattern recognition sensitivity to optimize both productivity and accuracy. By changing these parameters based on the specific event stream characteristics, the system maintains rule accuracy while achieving automated rule generation.
Data Source
AI summary
A novel and useful mechanism enabling a standard learning algorithm to generate rules for complex event processing (CEP) systems. The method creates rules that infer previously defined output events by creating input event feature vectors for each targeted output event. In addition, a method for automatically generating CEP system rules to infer output events which are anomalies (i.e. statistical outliers) of input event sequences is disclosed. Input feature vectors consisting of multiple input events and parameters for each targeted output event are then input into a standard learning algorithm to generate CEP system rules.