Benchmark
SummaC
A benchmark for evaluating the factual consistency of summaries, which judges the consistency between a summary and the source text based on sentence-level entailment relations.
Timeline (last 90 days)
Oct 23
Common automatic evaluation metrics include entailment scores based on NLI models, triple matching based on entity and relation extraction, and benchmarks such as FactCC and SummaC.
Unverified50%