Scalable Mining of Temporally Correlated Events
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network operations centers (NOCs) face challenges in managing and prioritizing a large volume of temporally correlated events from various sources, making it difficult to identify and act on related problems due to the complexity of combining and presenting multiple event relationships effectively.
Innovation Solution
A method involving the recording of event occurrences as bitmaps at fine and coarse-grained intervals, filtering events based on thresholds, grouping similar events, removing duplicates, creating transactions, applying optimized parallel frequent pattern growth, merging results, and representing the output as a distributed graph to identify and present temporally correlated events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If multiple separate sets of event relationships are extracted to provide additional useful information, then the information completeness is improved, but the system complexity and difficulty of combining relationships increases
Solution Approach 1:
The patent merges multiple separate sets of event relationships into a unified event graph structure. Different relationship types (temporal, topological, causal) are integrated into a single graph where nodes represent events and edges represent relationships, allowing comprehensive information to be presented in a unified manner rather than as separate disconnected sets.
Solution Approach 2:
The event graph serves as a universal data structure that can represent multiple types of relationships simultaneously. The graph structure is multi-functional, accommodating temporal sequences, topological connections, and causal relationships within the same framework, eliminating the need for separate structures for each relationship type.
2Loss of information
If multiple separate sets of event relationships are extracted to provide additional useful information, then the information completeness is improved, but the ease of operation for presenting relationships to users deteriorates
Solution Approach 1:
The patent merges multiple separate sets of event relationships into a unified event graph structure. Different relationship types (temporal, topological, causal) are integrated into a single graph where nodes represent events and edges represent relationships, allowing comprehensive information to be presented in a unified manner rather than as separate disconnected sets.
Solution Approach 2:
The patent transforms multiple flat relationship sets into a hierarchical graph structure with events as nodes and relationships as edges. This dimensional transformation organizes relationships in a multi-layered structure that is more intuitive for users to navigate and understand, moving from二维 tables to a structured graph representation.
3Measurement precision
If event data is processed in detail to identify temporally correlated events, then the measurement precision is improved, but the processing time increases
Solution Approach 1:
The patent segments the event processing task into distinct phases: event detection, relationship extraction, graph construction, and pattern mining. This segmentation allows each phase to be optimized independently, improving overall efficiency while maintaining precision in identifying temporally correlated events.
Solution Approach 2:
The patent performs preliminary filtering and grouping of events before detailed correlation analysis. Events are pre-processed to identify candidate sets based on temporal proximity and other criteria, reducing the search space for subsequent correlation detection and minimizing processing time while maintaining detection accuracy.
Data Source
AI summary
Scalable mining method, computer system and computer program product for temporally correlated events that works with input data that is non-transactional in nature and detects high correlation relationships between events with low support thresholds. The scalable mining method manages a large scale of input data by executing the mining process in a distributed manner in a clustered cloud environment via nodes.