Hierarchical Event Clustering for Root Cause Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In dynamic environments like telecommunications networks, identifying root causes of issues such as systematic errors and configuration problems in customer event sequences is challenging due to their unique and complex nature, leading to poor customer experience.
Innovation Solution
A system and method that groups customer event sequences into a hierarchical framework of patterns using event clustering, sequence clustering, and temporal clustering, followed by root-cause analysis to identify dominant patterns and significant sub-sequences, enabling effective root-cause analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If customer event sequences are analyzed individually to maintain unique details, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the complex analysis task into multiple hierarchical levels: first identifying individual event sequences, then grouping them into pattern clusters, and finally analyzing dominant patterns. This segmentation allows detailed individual analysis while managing overall system complexity through structured organization.
Solution Approach 2:
The patent introduces a hierarchical dimension to the analysis, moving from individual sequence level to pattern cluster level to dominant pattern level. This dimensional transformation enables the system to handle complexity by distributing analysis across multiple levels rather than attempting to analyze all sequences simultaneously at one level.
2Reliability
If all customer event sequences are analyzed in detail to identify root causes, then reliability is improved, but loss of time increases
Solution Approach 1:
The patent applies partial action by focusing analysis on dominant patterns that represent the most frequent or significant event sequences, rather than analyzing every single customer event sequence in equal detail. This allows the system to identify root causes with high reliability by concentrating on the most relevant patterns while reducing overall analysis time.
Solution Approach 2:
The patent performs preliminary grouping and clustering of event sequences into patterns before conducting detailed root-cause analysis. This preliminary organization identifies which sequences are similar and can be analyzed together, reducing the total number of detailed analyses required while maintaining comprehensive coverage.
3Device complexity
If event sequences are grouped into patterns to reduce complexity, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent applies local quality by maintaining different levels of detail at different hierarchical levels: individual sequences retain full detail for accurate pattern matching, while pattern clusters provide summarized views for high-level analysis. This allows the system to reduce overall complexity while preserving measurement precision where it matters most - in the detailed sequence-level analysis.
Data Source
AI summary
A system, method, and computer program product are provided for identification of common root causes with sequential patterns. In use, a plurality of customer event sequences are identified. Additionally, the plurality of customer event sequences are grouped into a hierarchical framework of patterns. Further, a root-cause analysis (RCA) is performed for the customer event sequences utilizing the hierarchical framework of patterns.