Automated Bug Management Tool for Software Localization Noise Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for testing and managing bugs in software user interfaces are time-consuming and costly, especially during product development and localization, as they often result in false positives and require extensive manual review across multiple language versions.
Innovation Solution
An automated bug management method that filters out noise bugs by comparing potential bugs in a new software version with previously reviewed bugs in a baseline version, using language-insensitive references and severity levels to identify and remove falsely identified issues, thereby reducing the number of bugs that need to be reviewed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If automated bug detection is performed across all language versions, then bug detection coverage is improved, but false positives increase and review time increases
Solution Approach 1:
The system performs preliminary bug detection on the source language version before localization, stores these detected bugs, and then uses them as a reference to filter bugs in translated versions. This preliminary action prevents re-detection of the same bugs in different language versions, reducing review time while maintaining detection coverage.
Solution Approach 2:
The system creates a copy of the bug data from the source language version and uses it to compare against bugs detected in translated versions. By copying and reusing the bug detection results across language versions, the system avoids redundant detection and review of identical UI defects in different languages.
2Measurement precision
If full UI testing is performed in each new build, then defect detection is improved, but testing cost and time increase
Solution Approach 1:
The system performs UI defect detection once on the source language version as a preliminary step, then reuses these detection results when testing translated versions. This eliminates the need to perform identical full UI testing in each language version, improving testing efficiency while maintaining defect detection accuracy through the comparison mechanism.
Solution Approach 2:
The bug detection mechanism is designed to be universal across multiple language versions. The same detection rules and criteria applied to the source language version are automatically applied to all translated versions, allowing a single testing effort to serve multiple language versions simultaneously.
3Reliability
If bugs are reviewed in each language version independently, then language-specific bugs are caught, but redundant review of identical bugs occurs
Solution Approach 1:
The system copies bug information from the source language version and compares it with bugs found in translated versions. When identical UI defects are detected in different language versions, the system recognizes them as duplicates through comparison and prevents redundant review, while still maintaining the ability to detect language-specific bugs that differ from the source.
Solution Approach 2:
The system implements a feedback mechanism where bug review results from one language version feed into the review process of subsequent language versions. By comparing detected bugs against previously reviewed bugs, the system provides feedback that eliminates redundant reviews while preserving language-specific defect detection.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An automated bug management method and a bug management tool are described. The method filters data received which identifies one or more potential bugs in the user interface of a version of a software product. This is achieved by accessing data relating to previously reviewed bugs in a user interface of another version of the software product and comparing the potential bugs with the reviewed bugs to determine whether any of the potential bugs are noise. The filtering process removes data relating to those potential bugs which are noise and then the filtered data is used to update a store of bug data.