| 研究課題/領域番号 |
22K02874
|
|---|
| 研究機関 | 京都先端科学大学 |
|---|
研究代表者 |
マルチュケ モリツ 京都先端科学大学, 経済経営学部, 准教授 (80738584)
|
|---|
| 研究分担者 |
林 勇吾 立命館大学, 総合心理学部, 教授 (60437085)
|
|---|
| 研究期間 (年度) |
2022-04-01 – 2026-03-31
|
| キーワード | E-Learning / Online Course / Kano model / NLP |
|---|
| 研究実績の概要 |
I have (co-)authored three journal papers (all first author, including a high-impact journal Wiley Disasters), four conference papers (two first author), and one book (organizing committee member for proceedings of ACM IC4E). The knowledge gained from collaborating with international researchers in various fields has prompted future research ideas, which are currently implemented. These include the extension of the current research topic to the use and analysis of generative AI and large language models (LLMs) in e-learning settings.
The research system setup was updated (dynamic CMS website with necessary functionality and questionnaire updates), to include generative AI and natural language processing (NLP).
A new approach was implemented by using dynamic topic modeling to classify and label e-learning comments to analyze topics and their sentiment. Findings were published in IEEE WAIE 2024 proceedings.
I have extended my research collaborative community to include researchers from Brawijaya University in Indonesia (e-learning and IoT experts), Munich University of Applied Sciences in Germany (LLM expert and PhD student collaboration), International Balkan University in North Macedonia (Psychology and Psycho-technology expert), as well as an international team at Ritsumeikan University and Kyoto University of Advanced Science (technology and AI experts, international economy experts). This research grant and research project had a wide-reaching cross-pollination effect of e-learning with multiple research fields.
|
| 現在までの達成度 |
現在までの達成度
2: おおむね順調に進展している
理由
While there was an initial delay due to a job change in AY2023 and the project was extended by one year, class and analysis are being pursued at a smooth rate now.
|
| 今後の研究の推進方策 |
Data from further courses is being implemented and evaluated. Data gathered for each course has been extended to track students interaction and satisfaction with course content as well as a chatbot feature for students to interact with the course content.
Results from additional classes will be consolidated with previous findings. Advanced language models (transformer based, e.g., BERT or LAMA), including Topic Modeling, UMAP clustering, etc. will be used to evaluate freeform text comments and create a hybrid model with the Kano method. New and updated results will be published in top international conferences. Aggregate results are planned to be published in renowned journals.
|
| 次年度使用額が生じた理由 |
The remaining funds is planned to be used for website and online course hosting and plugin costs, purchasing teaching hardware, and covering publication fees.
|
