Enhancing Machine Translation Quality Estimation via Fine-Grained Error Analysis and Large Language Model

Keyphrases

• Quality Estimation 100%
• Error Analysis 100%
• Machine Translation Evaluation 100%
• Large Language Models 100%
• Leveling Errors 66%
• Word-level 66%
• Sentence-level 66%
• Span Detection 66%
• Performance Enhancement 33%
• Performance Estimation 33%
• Optimal Performance 33%
• Detector Response 33%
• Error Detection 33%
• Classification Task 33%
• Enhanced Model 33%
• Machine Translation 33%
• Special Segment 33%
• Model Output 33%
• Language Pair 33%
• Chinese English 33%
• Severity Assessment 33%
• Fine-tuning Strategy 33%
• Span Extraction 33%
• Prompt Tuning 33%
• Task Granularity 33%
• Fine-tuned Model 33%

Computer Science

• Machine Translation 100%
• Large Language Model 100%
• Granularity 50%
• Optimal Performance 50%
• Error Detection 50%
• Classification Task 50%
• Estimation Performance 50%
• Enhance Performance 50%

Psychology

• Optimal Performance 100%