User Translation Feedback Integration Through Glossary Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual and machine translation of software applications often result in incorrectly translated words due to polysemous words, making it difficult to determine which instances of a word should be updated when a translation error is identified, and this issue is exacerbated by the lack of context in user interfaces.
Innovation Solution
A system and method using clustering and association analysis to determine which instances of a translated word to update based on user feedback, employing a glossary relationship set to group words by functional use cases and apply association relationship analysis for accurate translation adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual or machine translation is used for software applications, then translation coverage is achieved, but translation accuracy deteriorates due to polysemous words and lack of context
Solution Approach 1:
The patent introduces an intermediary system that acts as a mediator between the translated text and the final output. This intermediary performs consistency checks and context verification to ensure accurate translation of polysemous words, resolving the contradiction between achieving translation coverage and maintaining translation accuracy.
Solution Approach 2:
The patent implements a feedback mechanism where translation results are verified against the original context and glossary relationships. This feedback loop allows the system to identify and correct translation errors, particularly for polysemous words, thereby improving translation accuracy while maintaining comprehensive translation coverage.
2Stability of the object's composition
If all instances of a word are updated when a translation error is identified, then translation consistency is improved, but incorrect updates occur due to polysemous words with different meanings
Solution Approach 1:
The patent applies local quality by making different parts of the translation system have different functions. Specifically, it uses glossary relationships to identify which instances of a word should be updated together, rather than updating all instances uniformly. This allows the system to maintain translation consistency for related terms while avoiding incorrect updates of polysemous words with different meanings.
Solution Approach 2:
The patent segments the translation update process by dividing word instances into groups based on their glossary relationships. This segmentation allows the system to update only the appropriate instances of a word when a translation error is identified, preventing both over-update and under-update scenarios.
3Reliability
If user feedback is manually processed to correct translations, then translation accuracy can be improved, but processing time and resources increase significantly
Solution Approach 1:
The patent enables the translation system to self-correct by automatically processing user feedback through algorithmic consistency checks and glossary relationship analysis. This self-service capability allows the system to improve translation accuracy without requiring extensive manual intervention, thereby reducing feedback processing time and resource consumption.
Solution Approach 2:
The patent performs preliminary actions by pre-establishing glossary relationships and translation memory before the actual translation and feedback processing occurs. This preparation allows the system to quickly and efficiently process user feedback without requiring extensive computational resources during the feedback processing stage.
Data Source
AI summary
A method includes: receiving, by a computing device, user input indicating an incorrect translation of a word appearing in an interface of an application; identifying, by the computing device, other instances of the word in other interfaces of the application, wherein the identifying is performed using a glossary relationship set that is based on association analysis, wherein the other instances of the word constitute less than all instances of the word appearing in all interfaces of the application; and generating, by the computing device, a new version of the application having a revised translation of the word in the interface and the other instances of the word in the other interfaces.