Metric & Benchmark
ROUGE
Recall-oriented n-gram and longest-common-subsequence metric for summarisation eval.
Definition
ROUGE (Recall-Oriented Understudy for Gisting Evaluation) measures overlap between generated and reference summaries using n-grams (ROUGE-N), longest common subsequence (ROUGE-L), or skip-bigrams (ROUGE-S). It is the de facto summarisation benchmark scorer.
Common use cases
- Summarisation
- Headline generation
- Paraphrasing