論文 Development of LLM-assisted data curation tools for the Starrydata materials science database

Yukari Katsura SAMURAI ORCID (National Institute for Materials Science) ; Tomoya Mato SAMURAI ORCID (Center for Basic Research on Materials, National Institute for Materials Science) ; Yu Takada SAMURAI ORCID (Center for Basic Research on Materials, National Institute for Materials Science) ; Eiji Koyama (Center for Basic Research on Materials, National Institute for Materials Science (NIMS)) ; Dewi Yana (Center for Basic Research on Materials, National Institute for Materials Science (NIMS)) ; Atsumi Tanaka (Center for Basic Research on Materials, National Institute for Materials Science (NIMS)) ; Masaya Kumagai (Sakura Internet Research Center, Sakura Internet Inc.)

コレクション

引用
Yukari Katsura, Tomoya Mato, Yu Takada, Eiji Koyama, Dewi Yana, Atsumi Tanaka, Masaya Kumagai. Development of LLM-assisted data curation tools for the Starrydata materials science database. Science and Technology of Advanced Materials: Methods. 2025, 5 (1), 2590811. https://doi.org/10.1080/27660400.2025.2590811

説明:

(abstract)

We developed two LLM-assisted tools to accelerate data collection from materials science publications for the Starrydata database. The first tool, Starrydata Auto-Suggestion, generates concise English descriptions from abstracts and methods that conform to our database schema, integrated into the Starrydata2 platform using lightweight models. The second tool is a dual-component system: Auto-Summary GPT processes PDF files to generate comprehensive JSON output capturing all figures, tables, and samples, while Auto-Summary Viewer transforms this into interactive tables for efficient curator review. These tools enhance curation efficiency and advance automated scientific database construction.

権利情報:

キーワード: Materials informatics, materials database, data curation, literature data mining, automated data extraction, automated knowledge extraction, large language model

刊行年月日: 2025-12-31

出版者: Informa UK Limited

掲載誌:

  • Science and Technology of Advanced Materials: Methods (ISSN: 27660400) vol. 5 issue. 1 2590811

研究助成金:

  • JST-CREST Grant JPMJCR19J1
  • Kazuchika Okura Memorial Foundation

原稿種別: 出版者版 (Version of record)

MDR DOI:

公開URL: https://doi.org/10.1080/27660400.2025.2590811

関連資料:

その他の識別子:

連絡先:

更新時刻: 2026-01-20 09:35:27 +0900

MDRでの公開時刻: 2026-01-20 12:23:04 +0900

ファイル名 サイズ
ファイル名 Development of LLM-assisted data curation tools for the Starrydata materials science database.pdf (サムネイル)
application/pdf
サイズ 12.5MB 詳細