試験
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 34 / 65
Which metric is commonly used to quantitatively evaluate the lexical diversity of generated summaries?
①ROUGE score
②Distinct-n (the ratio of unique n-grams that appear)
③Type-Token Ratio (vocabulary diversity ratio)
④Perplexity (prediction difficulty of language models)
⑤BLEU score