Dynamic Translation Re-ranking via User Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine translation systems require human intervention to correct errors, which is inefficient and relies on presenting alternative translations to human translators for selection, but lacks a method to dynamically re-rank alternatives based on user input.
Innovation Solution
A system and method that presents machine translations along with alternative translations, where user selection of an alternative re-ranks the remaining options, allowing continuous proofing and improvement until the user confirms completion, utilizing a computing device to manage and re-order translation alternatives based on user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine translation is used to reduce translation time, then productivity increases, but translation accuracy deteriorates due to possible incorrect translations
Solution Approach 1:
The patent introduces an intermediary system between machine translation and final output that presents multiple alternative translations to users. This intermediary layer allows users to select or combine translations, thereby maintaining the speed advantage of machine translation while improving accuracy through human judgment and alternative options.
Solution Approach 2:
The system changes the parameter of translation presentation by providing multiple alternative translations with different quality levels or confidence scores. Users can adjust their selection based on these parameters, allowing them to balance speed and accuracy according to their needs for each specific translation task.
2Reliability
If human translators are employed to correct machine translation errors, then translation accuracy improves, but productivity decreases due to increased time requirements
Solution Approach 1:
Instead of requiring human translators to review and correct entire translations, the system presents only alternative translations for specific portions where uncertainty exists. Users perform partial action by selecting from alternatives rather than completely reviewing and rewriting translations, thereby maintaining accuracy improvements while reducing the time burden on human translators.
3Reliability
If alternative translation options are presented to users for selection, then translation accuracy improves, but device complexity increases due to the need to manage and re-rank multiple alternatives
Solution Approach 1:
The system performs preliminary action by pre-generating and ranking multiple alternative translations before user interaction. The alternatives are pre-processed and organized in a structured manner with confidence scores or quality metrics, reducing the complexity of real-time processing during user selection and making the interaction simpler rather than more complex.
4Reliability
If the system re-ranks alternative translations based on user selection, then translation quality improves through iterative learning, but loss of time increases due to the re-ranking process
Solution Approach 1:
The system implements feedback mechanisms where user selections are used to re-rank alternative translations for future presentations. This feedback loop allows the system to learn from user preferences and improve translation quality over time. The re-ranking process occurs in the background without significantly impacting user interaction time, as it prepares alternatives for subsequent queries rather than blocking current operations.
Data Source
AI summary
Disclosed herein are systems, methods, and non-transitory computer-readable storage media for presenting a machine translation and alternative translations to a user, where a selection of any particular alternative translation results in the re-ranking of the remaining alternatives. The system then presents these re-ranked alternatives to the user, who can continue proofing the machine translation using the re-ranked alternatives or by typing an improved translation. This process continues until the user indicates that the current portion of the translation is complete, at which point the system moves to the next portion.