Intelligent Customer Service Analysis System for Common Error Sequence Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In complex production environments, analyzing error logs from multiple sources to find the root cause of issues is challenging due to noisy information and the inability to leverage historic logs for future diagnostics, as existing methods focus on single log files and fail to detect event sequences.
Innovation Solution
A method and system for generating a metadata set from source information, including error log information that forms an error sequence, and creating a common error sequence set across multiple metadata sets using improved algorithms like the longest common subsequence (LCS') to identify causality and repeatability, with features like normalization, de-duplication, and filtering to reduce noise and enhance pattern recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing methods analyze a single log file of a single case, then the analysis process is simple, but it cannot detect event sequences and get useful information for root cause analysis
Solution Approach 1:
The patent merges multiple log files from different sources and time periods into a unified analysis system. It combines error logs, system logs, and application logs into a common sequence set, enabling comprehensive root cause analysis that leverages information across multiple cases rather than analyzing isolated single-case logs.
Solution Approach 2:
The patent transitions from analyzing single log files in isolation to analyzing log sequences across multiple dimensions including time sequences, error propagation chains, and cross-source correlations. This dimensional expansion enables detection of event sequences and causal relationships that are invisible in single-case analysis.
2Adaptability or versatility
If existing methods view single log files, then the analysis is straightforward, but historic logs cannot be leveraged for future diagnostic and serious problem prevention
Solution Approach 1:
The patent performs preliminary actions by pre-processing and storing log information in structured metadata formats with extracted features and sequences. This preparation enables rapid retrieval and analysis of historic logs when diagnostic needs arise, transforming raw historical data into actionable intelligence without requiring complex real-time processing during incident response.
Solution Approach 2:
The patent creates simplified copies of complex log data by extracting essential features, error sequences, and metadata representations. These copies preserve the diagnostic value of historic logs while reducing complexity, enabling efficient pattern matching and comparison across different time periods and systems.
3Measurement precision
If existing methods analyze complex production components without comparing different log information, then the analysis process is simple, but it is very hard to get the root cause from a complex issue
Solution Approach 1:
The patent implements feedback mechanisms by comparing current error sequences against patterns learned from historic logs and other system sources. This feedback loop enables the system to quickly identify known problem patterns and accelerate root cause identification, reducing analysis time while maintaining high accuracy through iterative pattern recognition.
Solution Approach 2:
The patent performs preliminary comparison and pattern matching by pre-processing logs into standardized sequences and storing them for rapid retrieval. When a new issue arises, the system immediately compares it against the pre-processed historical data, eliminating the need for time-consuming analysis from scratch and enabling fast root cause identification.
4Loss of information
If existing methods look for problems by viewing a single log file, then the method is simple to implement, but it cannot detect event sequences and get useful information
Solution Approach 1:
The patent segments log information into distinct metadata components including error sequences, system states, timestamps, and causal relationships. This segmentation preserves event sequence information by structuring it into analyzable units while managing complexity through organized, modular data representation that can be processed independently and recombined.
Data Source
AI summary
The present invention provides a method and system for information analysis. The method extracts a plurality of metadata from a source information set so as to generate a metadata set, the metadata comprising error log information that forms an error sequence in the metadata set; and generates a common error sequence set for a plurality of the metadata sets. By means of the method, it is possible to easily obtain a global error sequence pattern, and easily compare a new error with a previous error sequence pattern so as to prevent, diagnose and recover the new error.