[2O4-GS-13] AI application: Agriculture, forestry and fisheries

[2O4-GS-13-04] NARO Linked Database for Agricultural Research Data Management

Takahiro Kawamura¹, 〇Tetsuo Katsuragi¹, Motoko Inatomi¹, Hiromi Kanegae¹, Hisashi Eguchi¹ (1. NARO)

Keywords:Agricultural intelligence, Database, Agricultural Research Data

Recently, research data are strongly required to be effectively utilized for research acceleration and new knowledge
creation. Thus, national research institutes and universities are developing research data management systems, and
we, the National Institute of Agriculture, Forestry and Fisheries (NARO), constructed NARO Linked Database
(narolin DB) and started its operation in April 2020. In the narolin DB, various research data on agriculture
(genome, variety, pest, environmental information, and so forth) are cataloged with common metadata based on
the FAIR principle. On the other hand, the relationship between complicated data in natural science is described in
Resource Description Framework (RDF), Property Graph, or RDB format, and it provides mechanisms to facilitate
the application of statistical analysis and machine learning. In particular, in order to connect research data of long
research periods such as agriculture and environment to cross-sectoral research in a faster cycle, daily-used shared
folders and a research data repository have been integrated into a narolin DB. In this paper, after outlining the
trend of the research data repository, the system architecture of the narolin DB and the NARO common metadata
schema are introduced, and finally, we conclude this paper with future issues.