| Presentation | 2021-02-18 Classification of disaster tweets for damage assessment, and improvement by feature analysis Yuto Oikawa, Ptaszynski Michal, Fumito Masui, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In extracting tweets useful in rescue missions during disasters, previous research have focused on extracting tweets containing specific addresses or locations. We assume that tweets without addresses can also be useful for disaster relief as the location can be inferred or written indirectly. In this study, we focus on extracting tweets from users who directly experienced the disaster (tweets with high directness) and classified them into three classes using BERT, based on the assumption that the tweets, when provided to rescue teams, can be useful for evaluating the disaster situation. Additionally, we performed feature analysis of the training data, which helped us update the annotation criteria, and improve the classification efficacy. The results were satisfying enough to be considered for application in efficient information extraction during disasters. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Disaster Information / Rescue Request / Twitter / BERT / Document Classification / Text Mining |
| Paper # | NLC2020-22 |
| Date of Issue | 2021-02-11 (NLC) |
| Conference Information | |
| Committee | NLC |
|---|---|
| Conference Date | 2021/2/18(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | The 17th Text Analytics Symposium |
| Chair | Kazutaka Shimada(Kyushu Inst. of Tech.) |
| Vice Chair | Mitsuo Yoshida(Toyohashi Univ. of Tech.) / 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) | Classification of disaster tweets for damage assessment, and improvement by feature analysis |
| Sub Title (in English) | |
| Keyword(1) | Disaster Information |
| Keyword(2) | Rescue Request |
| Keyword(3) | |
| Keyword(4) | BERT |
| Keyword(5) | Document Classification |
| Keyword(6) | Text Mining |
| 1st Author's Name | Yuto Oikawa |
| 1st Author's Affiliation | Kitami Institute of Technology(KIT) |
| 2nd Author's Name | Ptaszynski Michal |
| 2nd Author's Affiliation | Kitami Institute of Technology(KIT) |
| 3rd Author's Name | Fumito Masui |
| 3rd Author's Affiliation | Kitami Institute of Technology(KIT) |
| Date | 2021-02-18 |
| Paper # | NLC2020-22 |
| Volume (vol) | vol.120 |
| Number (no) | NLC-374 |
| Page | pp.pp.7-12(NLC), |
| #Pages | 6 |
| Date of Issue | 2021-02-11 (NLC) |