Entropy-Based Source File Noise Reduction for Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current code testing methods are inconsistent and often result in poor testing coverage, leading to frequent post-deployment failures and defects due to inadequate scanning that fails to account for unique organizational factors, resulting in numerous false positives.
Innovation Solution
A method that scans source files to determine entropy values, acquires and adjusts threshold values based on file type and organizational specifics, removes high and low entropy strings, and executes targeted tests to reduce noise and false positives, using a processor and memory to ensure accurate testing before deployment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional scanning with generic high threshold is used, then false positives are reduced, but testing coverage becomes insufficient and post-deployment failures increase
Solution Approach 1:
The patent applies local quality by implementing organization-specific threshold values tailored to different file types and projects, rather than using a generic high threshold universally. This allows the scanning system to maintain high precision in reducing false positives while adapting sensitivity levels to local characteristics of different codebases, thereby preserving testing coverage without excessive false positives
Solution Approach 2:
The patent implements dynamic threshold adjustment based on entropy values calculated from actual code patterns. The system dynamically adapts threshold levels according to the specific organizational context and file type being scanned, enabling it to maintain optimal balance between false positive reduction and testing coverage across diverse scenarios
2Measurement precision
If generic high threshold scanning is applied, then false positives decrease, but organizational unique factors are not accounted for leading to inconsistent testing
Solution Approach 1:
The system implements local quality by creating organization-specific and file-type-specific threshold configurations. Each organization can define their own threshold values based on their unique code patterns, project requirements, and risk tolerance, making the scanning adaptable to local characteristics rather than applying a one-size-fits-all approach
Solution Approach 2:
The patent changes the parameter of threshold values from fixed generic highs to dynamic organization-specific values. The system calculates entropy-based thresholds that adapt to organizational patterns, allowing the same scanning mechanism to work effectively across different organizations with their unique factors
3Reliability
If extensive testing is performed on all strings, then testing coverage is improved, but noise and resource consumption increase
Solution Approach 1:
The patent applies the extraction principle by removing high-entropy strings (likely non-codified data, comments, or irrelevant content) from the testing scope before execution. This reduces the volume of data subjected to testing while maintaining coverage of critical low-entropy code patterns, thereby decreasing noise and resource consumption without sacrificing essential testing coverage
Data Source
AI summary
A method for performing reducing noise and false positive event detection in testing is disclosed. The method includes performing scanning of a source file including multiple strings; determining, an entropy value of each string included in the multiple strings; acquiring, from a central repository, a first threshold value; identifying one or more strings having an entropy value higher than the first threshold value; removing the one or more strings having the entropy value higher than the first threshold value from the source file; deriving a second threshold value based on the determined entropy values and the first threshold value; identifying one or more strings having an entropy value lower than the second threshold value; removing the one or more strings having the entropy value lower than the second threshold value from the source file; and executing one or more tests on strings remaining in the source file.