EXTENDED ABSTRACT: Lots of distributed materials data repositories and platforms have been established and used in data-driven material discovery and design. To facilitate more efficient and systematic data-driven materials research, data linking and sharing between these distributed data infrastructures must be essential. In this presentation, some technical issues for this data sharing were addressed and strategical approaches to them were discussed. Data diversity generated from multiple data sources and repositories requires standards for reading and understanding the data, and standards including data common format and vocabularies were being built up. Basically, it was proposed that materials data structure would be presented with three elements of value-uncertainty-measurement, and the respective elements could be represented with a common format. Data vocabularies relating to structure-process-property-performance of materials, which relationship can be generally described to material, were being standardized. In the K-MDS platform, the data vocabularies were systematically classified and managed with 'Materials System'. To integrate or combine the data from multiple heterogeneous sources,
ontology-based database is required. So, material specific-ontology development has been tried and being supported. Some reference templates for materials and materials systems data comparable to material-specific ontologies were developed, and, currently, K-MDS platform is serving 15 reference data templates. Data collection(registration) tools optimized with data schema of 'Materials System' was developed and implemented into the K-MDS platform. Data mining techniques based on BERT and GPT also were being developed.
Keywords: Materials data platform(repositories); Data sharing; Materials Data Standard; K-MDS platform;
Acknowledgement:
This research was supported by the National Center for Materials Research Data (NCMRD) through the National Research
Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF-2021M3A7C2089748)
Dr. Yong-Hak Huh is a principal research scientist at Korea Research Institute of Standards and Science (KRISS) and has served as Head of National Center for Materials Research Data since Oct., 2021. He received his Ph. D from Korea Advanced Institute of Science and Technology (KAIST) in 1998. He joined KRISS in 1987 and contributed to the development of materials standards. He has served as a project leader and convenor in ISO(TC164), IEC(TC47) and APMP(TCMM) to establish international test standards for material properties. He has published about 100 research papers and served as an editorial board member of ISIJ.