|
ヤマモト エイコ
Eiko Yamamoto
山本 英子 所属 経済情報学部 職種 教授 |
|
| 言語種別 | 英語 |
| 発行・発表の年月 | 2020/10 |
| 形態種別 | 研究論文（国際会議プロシーディングス） |
| 査読 | 査読あり |
| 標題 | Improving Semantic Similarity Calculation of Japanese Text for MT Evaluation |
| 執筆形態 | 共著 |
| 掲載誌名 | Proc. of 34th Pacific Asia Conference on Language, Information and Computation (PACLIC2020) |
| 掲載区分 | 国外 |
| 総ページ数 | 9 |
| 著者・共著者 | Yuki Tanahashi, Kyoko Kanzaki, Eiko Yamamoto and Hitoshi Isahara |
| 概要 | We verified the method by calculating the Pearson correlation between the modified BERTScore and human-rated scores. Further-more, we set four BERT models and two kinds of corpora to calculate idf value, and investigated which setting is most suitable for evaluation of novel translation. As a result, the setting with the model based on novel corpus, the idf based novel corpus and the penalty had the highest correlation with human-rated scores. |