Context-Aware Software Translation via Test Step GUI Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current software translation processes suffer from poor quality translations due to a lack of context, leading to increased costs and resource expenditure, as human translators unfamiliar with the application's GUI struggle to provide accurate translations, and machine-translation services fail to improve translation algorithms effectively.
Innovation Solution
A system and method that utilize test steps and subject application displays to determine machine-translations, obtain user-translations, and update translation scripts, allowing translators to edit GUI strings while executing the application, thereby improving translation quality and efficiency by providing context-aware translations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If human translators translate application strings without context, then translation process is simple, but translation quality deteriorates
Solution Approach 1:
The system performs preliminary action by executing the subject application to generate test steps and capture display states before translation occurs. This allows translators to receive strings with contextual information about their usage scenarios, enabling accurate translations while maintaining process simplicity.
Solution Approach 2:
The system introduces an intermediary mechanism that captures application state information (test steps, display states, UI elements) and transmits it to translators. This intermediary layer provides context without requiring translators to directly analyze complex application code, resolving the contradiction between simplicity and quality.
2Productivity
If translators are separated from application development, then development efficiency is improved, but translation accuracy deteriorates
Solution Approach 1:
The system creates a copy of the application execution environment that generates test steps and display states. Translators receive this contextual copy information without needing to access the actual application code or execution environment, maintaining separation while improving accuracy through contextual copying of usage scenarios.
Solution Approach 2:
The system introduces an intermediary layer that bridges translators and developers by capturing application context information independently. This intermediary mechanism allows translators to work with contextualized strings without being part of the development team, resolving the contradiction between productivity and accuracy.
3Speed
If machine-translation is used without context, then translation speed is improved, but translation quality deteriorates
Solution Approach 1:
The system performs preliminary capture of application context information (test steps, display states, UI element relationships) before machine translation occurs. This contextual data is then provided to machine translation services, enabling high-speed automated translation with improved accuracy through contextual awareness.
Solution Approach 2:
The system changes the parameters fed to machine translation services by adding contextual information about application state, usage scenarios, and UI context. This parameter enhancement allows machine translation to operate at high speed while producing higher quality translations through enriched input data.
4Adaptability or versatility
If translators update machine-translation scripts after development, then development flexibility is improved, but translation cost increases
Solution Approach 1:
The system performs preliminary capture of application context and generation of test steps during development. This allows translation resources to be prepared and updated in advance or continuously, enabling flexible development while reducing the need for costly re-translation by capturing context early when changes occur.
Solution Approach 2:
The system establishes feedback loops where application execution context is continuously captured and fed back to translation processes. This feedback mechanism enables automatic updates of translation resources as the application evolves, maintaining flexibility while reducing costs through continuous improvement rather than periodic re-translation.
Data Source
AI summary
In one example, a processor may: execute a machine-translation script to generate a machine-translation for a set of strings to be displayed upon execution of a subject application; cause a first display including a listing of testing actions to be performed by a test application; cause a second display that includes a GUI of the subject application, the second display including the set of strings; receive a user-translation for each of the strings via the GUI; and update the machine-translation script to include the received user-translations.