Pattern-Based Rule Extraction from Test Log Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Compiling problem diagnostic knowledge in application development is a reactive, tedious, and largely manual process, requiring significant expertise and time to identify, resolve, and document problems and solutions across various scenarios.
Innovation Solution
A method involving the creation of pattern-based rules from log entries associated with test cases, which are stored in a knowledge database, allowing for the generation of messages containing stored knowledge when a partial match is identified between program activity patterns, facilitating faster problem identification and resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If problem diagnostic knowledge is compiled manually in response to user-reported problems, then diagnostic knowledge can be created with expertise, but the process is reactive, tedious, and consumes significant time and energy
Solution Approach 1:
The system performs preliminary action by automatically generating problem diagnostic knowledge from test case log entries before users encounter and report problems. The rule generator analyzes log data from test cases and creates pattern-based rules in advance, so when a user encounters an issue, the diagnostic knowledge is already available and can be immediately applied, eliminating the reactive and time-consuming manual compilation process
Solution Approach 2:
The system uses copying by creating pattern-based rules that replicate the diagnostic patterns found in test case log entries. Instead of manually documenting each problem scenario, the system extracts patterns from log data and creates reusable rule templates that can be applied to identify similar problems automatically, significantly reducing the time and effort required for knowledge compilation while maintaining accuracy
2Loss of information
If manual documentation of problems and solutions is performed across expansive scenarios, then comprehensive diagnostic knowledge can be accumulated, but the process requires significant expertise and investments of time and energy
Solution Approach 1:
The system creates comprehensive diagnostic knowledge coverage by copying and generalizing patterns from test case log entries into reusable rules. Instead of manually documenting each individual problem scenario, the system extracts underlying patterns that represent multiple scenarios, creating a compact yet comprehensive knowledge base that covers expansive scenarios without requiring proportional manual effort
Solution Approach 2:
The system applies parameter changes by transforming detailed log entry data into generalized pattern rules. The rule generator analyzes log data and converts specific instance details into parameterized patterns that can match multiple scenarios, thereby accumulating comprehensive diagnostic knowledge across expansive scenarios while improving creation efficiency by working with generalized representations rather than individual cases
3Extent of automation
If pattern-based rules are created from test case log entries, then knowledge extraction from testing scenarios is automated, but the system must process and analyze large volumes of log data
Solution Approach 1:
The system applies extraction by isolating and focusing on specific pattern elements within log data that are most relevant for diagnostic purposes. The rule generator extracts key patterns from test case log entries and creates rules based on these essential elements, automating knowledge extraction while managing data processing complexity by concentrating on the most diagnostic features rather than processing all log data in exhaustive detail
Solution Approach 2:
The system uses local quality by creating pattern rules that focus on specific local characteristics of log entries rather than requiring analysis of entire log datasets. The rule generator identifies and extracts localized patterns from log data that are sufficient for diagnostic purposes, enabling automation of knowledge extraction while reducing processing complexity by working with targeted local features rather than global data analysis
4Reliability
If reactive manual problem resolution is performed, then problems can be addressed with expert judgment, but the process is tedious and time-consuming for each individual problem
Solution Approach 1:
The system uses copying by creating pattern-based rules that replicate expert problem-resolution patterns extracted from test cases. When a user encounters a problem, the system matches the situation against stored patterns and automatically applies the corresponding resolution approach, maintaining the reliability of expert judgment while eliminating the time-consuming manual resolution process for each individual problem
Solution Approach 2:
The system performs preliminary action by pre-processing test case data to extract and store pattern-based resolution rules before actual problem occurrence. This allows the system to have diagnostic and resolution knowledge ready in advance, so when problems occur, they can be addressed quickly by matching against pre-prepared patterns rather than requiring reactive manual analysis, thereby maintaining accuracy while reducing resolution time
Data Source
AI summary
Embodiments of the present invention enable users to extract knowledge from testing scenarios performed during application development, and later employ that knowledge to interpret application usage scenarios to enhance serviceability of applications by expediting identification and solving of problems. In an exemplary embodiment, log data generated during simulation of test cases is analyzed to create one or more rules based on patterns in which one or more log entries appear in the log data. Later, log data may be analyzed to look for a pattern of log entries that matches a pattern-based rule, thereby facilitating faster identification and resolution of the problem.