Personalized Machine Translation via Online Feedback Adaptation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Machine translation systems lack personalization and improvement mechanisms, relying on static translation methodologies that do not adapt to user feedback, leading to suboptimal translation quality.

Innovation Solution

A personalized machine translation system that receives and processes translator feedback, using a feedback processor to classify useful feedback and incorporate it into the translation methodology, including dictionary updates and preference integration to enhance translation accuracy and context relevance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If translator feedback is incorporated into the translation methodology, then translation quality and personalization improve, but system complexity increases

Engineering Contradiction:

Improvetranslation qualityVSAvoidsystem complexity

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

The system implements a feedback mechanism where translator corrections and preferences are collected, processed, and incorporated back into the translation methodology. The feedback processor analyzes translator feedback and updates the translation model dynamically, creating a closed-loop system that continuously improves translation quality based on actual user interactions and corrections.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The translation methodology transitions from a static approach to a dynamic one that adapts in real-time based on translator feedback. The system dynamically updates translation preferences, terminology, and stylistic choices based on ongoing translator interactions, allowing the system to evolve and personalize translations for each user over time.

Inventive Principle:

Principle #15Dynamics

2Measurement precision

If translator feedback is processed and incorporated, then translation accuracy improves, but processing time and computational resources increase

Engineering Contradiction:

Improvetranslation accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary processing of translator feedback by categorizing and prioritizing corrections before full integration. High-impact corrections and recurring patterns are identified and applied first, allowing the system to quickly improve accuracy without processing every single feedback element in detail, thus reducing overall processing time.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system applies partial processing to feedback by focusing on the most significant corrections and patterns rather than uniformly processing all feedback equally. By identifying and prioritizing high-value feedback elements, the system achieves substantial accuracy improvements with reduced computational overhead compared to exhaustive processing of all feedback data.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS9152622B2Personalized machine translation via online adaptation

Publication Date: 2015.10.06 SDL INC

AI summary

Personalizing machine translation via online adaptation is described herein. According to some embodiments, methods for providing personalized machine translations may include receiving translator feedback regarding machine translations generated by a machine translation system for a translator, determining translator feedback that improves translations generated by the machine translation system, and incorporating the determined translator feedback into the translation methodology of the machine translation system to personalize the translation methodology.