Ties Matter: Meta-Evaluating Modern Metrics with Pairwise Accuracy and Tie Calibration

Keyphrases

• Metric Scores 100%
• Machine Translation 50%
• Ranking Method 50%
• Calibration Method 50%
• Fair Ranking 50%
• Fair Comparison 50%
• Meta-evaluation 50%
• Machine Translation Evaluation 50%
• Metric Performance 50%

Computer Science

• Machine Translation 100%
• Evaluation Metric 50%
• Performance Metric 50%
• Experimental Evidence 50%
• Fair Comparison 50%

Arts and Humanities

• Metrics 100%
• Variant 28%
• Automatic Translation 28%
• Literature 14%
• Weakness 14%
• Experimental 14%