| 研究課題/領域番号 |
21K17802
|
|---|
| 研究機関 | お茶の水女子大学 |
|---|
研究代表者 |
|
|---|
| 研究期間 (年度) |
2021-04-01 – 2024-03-31
|
| キーワード | NLP / adversarial training / machine learning |
|---|
| 研究実績の概要 |
The progress has been smooth so far. We were able to publish two papers at two international conferences last year. Following the original plan, we explored applying perturbations to other layers of the network during adversarial training, as well as the combination with multi-task learning.
The published papers are the following:
Lis Kanashiro Pereira, Kevin Duh, Fei Cheng, Masayuki Asahara, Ichiro Kobayashi . Attention-Focused Adversarial Training for Robust Temporal Reasoning. LREC 2022.
Lis Kanashiro Pereira, Ichiro Kobayashi. OCHADAI at SemEval-2022 Task 2: Adversarial Training for Multilingual Idiomaticity Detection. SemEval 2022.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
The progress has been smooth so far. Following the original plan, we are currently running further experiments on incorporating prior knowledge to adversarial training to generate better perturbations.
For information gathering, we are planning to attend the following upcoming conferences in June and July, respectively: JSAI 2023 and ACL 2023.
|
| 今後の研究の推進方策 |
We plan to evaluate our proposed model in the following months. If positive results (SOTA) are achieved, we will be submit them to the closest international conferences such EMNLP, COLING, etc. We also would like to share our progress in ANLP (言語処理学会年次大会) in March 2024.
|
| 次年度使用額が生じた理由 |
I plan to use the remaining amount to buy the necessary equipments such as GPU server machine, as well as for international and national conference attendance.
I also plan to use this budget to buy supplies such as books. I also want to use it for paid APIs , such as ChatGPT, which has become quite popular recently in the natural language processing field. We plan to use it in our research as well and further enhance our model's performance.
|
