Towards question-answering as an automatic metric for evaluating the content quality of a summary

Keyphrases

• Content Quality 100%
• Question Answering 100%
• Automatic Metrics 100%
• Reference Dependence 25%
• Benchmark Dataset 25%
• Token 25%
• Overlap Ratio 25%
• Evaluation Metrics 25%
• ROUGE 25%
• Performance Bottleneck 25%
• Information Overlap 25%
• Performance Estimate 25%
• State-of-the-art Models 25%

Computer Science

• Content Quality 100%
• Gold Standard 33%
• Desirable Property 33%
• Evaluation Metric 33%
• Performance Bound 33%
• Performance Bottleneck 33%