| Presentation | 2021-09-16 A causal relation extraction among distant texts using deep learning Pengju Gao, Tomohiro Yamasaki, Masahiro Ito, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Most of the Existing methods for causal relationship extraction utilize patterns such as clue expressions, but it is difficult to extract between those that do not fit the pattern or separated texts. Therefore, in our study, we tried an approach using the pre-trained model BERT. To overcome BERT input length limitation, we used Transformer to extract the causal relationships between the texts features. In addition, for data augmentation and imbalance controlling, we generated input from subsets of sentences instead of whole document. As a result, the extraction evaluation F value of the causal relationship improved from 0.400 to 0.458. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Causal relationship / Deep learning / BERT / Data augmentation |
| Paper # | NLC2021-8 |
| Date of Issue | 2021-09-09 (NLC) |
| Conference Information | |
| Committee | NLC |
|---|---|
| Conference Date | 2021/9/16(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | The 18th Text Analytics Symposium |
| Chair | Kazutaka Shimada(Kyushu Inst. of Tech.) |
| Vice Chair | Mitsuo Yoshida(Univ. of Tsukuba) / Takeshi Kobayakawa(NHK) |
| Secretary | Mitsuo Yoshida(Univ. of Tokyo) / Takeshi Kobayakawa(Hiroshima Univ. of Economics) |
| Assistant | Kanjin Takahashi(Sansan) / Ko Mitsuda(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Natural Language Understanding and Models of Communication |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A causal relation extraction among distant texts using deep learning |
| Sub Title (in English) | |
| Keyword(1) | Causal relationship |
| Keyword(2) | Deep learning |
| Keyword(3) | BERT |
| Keyword(4) | Data augmentation |
| 1st Author's Name | Pengju Gao |
| 1st Author's Affiliation | TOSHIBA Corporation(TOSHIBA) |
| 2nd Author's Name | Tomohiro Yamasaki |
| 2nd Author's Affiliation | TOSHIBA Corporation(TOSHIBA) |
| 3rd Author's Name | Masahiro Ito |
| 3rd Author's Affiliation | TOSHIBA Corporation(TOSHIBA) |
| Date | 2021-09-16 |
| Paper # | NLC2021-8 |
| Volume (vol) | vol.121 |
| Number (no) | NLC-178 |
| Page | pp.pp.11-16(NLC), |
| #Pages | 6 |
| Date of Issue | 2021-09-09 (NLC) |