Distance-Based Event Sequence Visualization for Rule Development
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Domain experts face challenges in developing event filtering rules due to the overwhelming number of events generated by complex systems, requiring efficient methods to identify and filter out uninformative events.
Innovation Solution
The method involves identifying sequences of events, calculating distance values, and visualizing them in a target metric space to group similar sequences together, allowing users to easily spot frequently occurring scenarios and segregate important events, thereby assisting in defining filtering rules.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If automated event filtering systems are used to reduce the number of events presented to users, then the information overload is reduced, but the accuracy and effectiveness of filtering depends heavily on expert knowledge and trial-and-error validation
Solution Approach 1:
The system automatically analyzes event sequences and generates filtering rules without requiring extensive manual expert intervention. The automated analysis engine processes event data, identifies patterns, and produces candidate filtering rules that can be validated and deployed with minimal human effort, enabling the system to serve itself in rule generation.
Solution Approach 2:
The system performs preliminary analysis of event sequences before final rule deployment. By pre-processing event data, identifying frequent patterns, and generating candidate rules in advance, the system prepares filtering rules that can be quickly validated and implemented, reducing the overall time required for rule development and deployment.
2Measurement precision
If expert knowledge is used to write and validate event filtering rules, then filtering accuracy is improved, but the process requires substantial trial and error and extensive domain knowledge
Solution Approach 1:
The system introduces an automated analysis engine as an intermediary between raw event data and filtering rules. This intermediary component performs pattern recognition, sequence analysis, and rule generation, bridging the gap between complex event data and usable filtering rules while reducing the need for direct expert intervention in the rule creation process.
Solution Approach 2:
The system implements feedback mechanisms where filtering rules are automatically validated against event data, and performance metrics are used to refine and improve rule quality. This iterative feedback process enables continuous improvement of filtering accuracy while reducing reliance on manual expert trial-and-error validation.
3Adaptability or versatility
If manual event filtering rule development is performed, then filtering rules can be customized to specific system needs, but the process is time-consuming and burdensome for operators
Solution Approach 1:
The system automatically generates customized filtering rules by analyzing system-specific event patterns and characteristics. The automated analysis engine adapts to the specific monitored system, identifying relevant event sequences and generating tailored filtering rules without requiring manual customization, thereby maintaining adaptability while dramatically improving productivity.
Solution Approach 2:
The system dynamically adjusts filtering rule parameters based on analyzed event patterns, system characteristics, and performance feedback. By automatically modifying rule parameters such as event sequence patterns, time windows, and filtering thresholds, the system customizes rules to specific system needs while maintaining high development efficiency through automated parameter optimization.
Data Source
AI summary
Event analysis methods and apparatus in which sequences (44, 46) of one or more events are identified based on event records (20) describing the events. Respective distance values (28) representing distances between ones of the sequences (44, 46) are determined. A configuration of points in a target metric space is constructed based on the distance values (28), where each of the points represents a respective one of the sequences (44, 46). A visual representation (38) of the configuration is presented on a display (34).