PoLyInfoと機械学習

近年、高分子分野においても、マテリアルズ・インフォマティクスへの関心が高まっている。一方で、高分子の多様性を反映して、系統的なデータの少なさや、データの統合が困難なことが問題になっている。ここでは、世界的な高分子データベースを俯瞰した上でのPoLyInfoの位置づけ、インフォマティクスの観点で見た特徴、関連の成果等をまとめる。

In recent years, there has been growing interest in materials informatics and machine learning in the field of polymers. On the other hand, reflecting the variety of polymers, the lack of systematic data and the difficulty of integrating data in a single format to cover such data have become a problem. The National Institute for Materials Science has collected polymer data from papers over a long period of time and published them in a database called "PoLyInfo". Here, we summarize the position of PoLyInfo in the global polymer database, the characteristics of PoLyInfo from the viewpoint of informatics, and related results. In particular, it will provide an overview of what is needed to extract the process-structure-property combination that is necessary for machine-readable representation of materials, with a view toward the future target of polymer informatics.