Development of Next-Generation Machine Intelligence for Predicting Material Properties, Considering the Influence of Experimental Processes and Sample Structures

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K22466
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 0401:Materials engineering, chemical engineering, and related fields
• Research Institution: Kyoto University
• Principal Investigator: Kumagai Masaya 京都大学, 複合原子力科学研究所, 特定助教 (00881054)
• Project Period (FY): 2020-09-11 – 2023-03-31
• Keywords: マテリアルズ・インフォマティクス / プロセス・インフォマティクス / 機械学習 / 材料工学

Outline of Final Research Achievements

In this study, we analyzed the relationships among process, structure, and physical properties and created our own large dataset. The process information in this study was collected by extracting text from PDFs of research papers. To analyze the relationship between structural information and physical properties, a machine learning model was constructed using X-ray diffraction patterns as input and crystal systems, volume, density, and volume modulus as learning targets. The original large dataset created during this research period has been publicly released on Figshare. The findings of this research were also disseminated externally through various means, including submissions to domestic and international conferences and journals.

Free Research Field

マテリアルズ・インフォマティクス

Academic Significance and Societal Importance of the Research Achievements

プロセス情報を含めた物性の予測を可能にすることは、新規材料の発見のみならず、製造プロセスの改善に貢献することができるため大きな意義がある。また、XRDと物性との関係性を大規模なデータを利用して明らかにできたことは、これまで結晶構造と物性の関係性を紐解く上で学術的に意義がある。さらに、本研究期間に作成した大規模実験データは、これからの実験MIを推進する基盤データとなると考えている。