Wataru Takahara of the Materials Informatics Laboratory received the JSAP Poster Award at the 73rd JSAP Spring Meeting 2026

Summary

In May 2026, Wataru Takahara of the Materials Informatics Laboratory at Nara Institute of Science and Technology received the JSAP Poster Award for his presentation at the 73^rd JSAP Spring Meeting 2026. He was a third-year doctoral student with a full-time job when the presentation was accepted and had become an alumnus by the time of the presentation. This award recognizes outstanding poster presentations delivered at the meeting.

JSAP Poster Award URL:

https://www.jsap.or.jp/poster-award

Presentation title

Domain-Specialized Method for Local LLMs Integrating Experimental Data and Scientific Literature — Proposal of MDSK-RAG —

Authors

Wataru Takahara^A, Yuichi Yamaguchi^B, Mai Ogano^B, Fuga Kakami^B, Yosuke Harashima^A, Tomoaki Takayama^A, Shogo Takasuka^A, Akihiko Kudo^B, Mikiya Fujii^A

A: NAIST, B: Tokyo University of Science

Research detail

This research proposes a framework for utilizing large language models (LLMs) in materials development by integrating experimental data and scientific literature on metal sulfide photocatalysts. By applying the proposed method to local LLMs, the framework enables domain specialization using relatively small datasets while maintaining the confidentiality of research data. It also allows continuous updating of knowledge sources and flexible extension without modifying the base model. This work demonstrates the potential of LLMs in materials development by combining experimental facts with knowledge from the scientific literature, even in limited-data environments.

Comment

I am deeply honored to receive the JSAP Poster Award at the 73rd JSAP Spring Meeting 2026. I would like to express my sincere gratitude to all of my collaborators for their invaluable guidance and support throughout this research. Related achievements from this work were also recognized at the 2026 MRS Spring Meeting & Exhibit, where it received the MRS MT01 Best Presentation Award. I am greatly encouraged that this research has been recognized both in Japan and overseas, and I will continue to contribute to the advancement of AI for Materials and AI for Science.