The construction of agricultural science data ontology network is an important part of agricultural science data analysis and mining. It can integrate data scattered in different databases and different formats, correlate and integrate data from different fields to form a more comprehensive data pool, support automatic analysis and mining of cross-domain and interdisciplinary data, and discover hidden knowledge, patterns and trends. In this paper, a database of 28 agricultural science data ontologies related to agriculture, crops, genes, and sequences is constructed, the storage standard of agricultural science data ontology is formulated, the agricultural science data ontology network based on the HugeGraph graph database is built, the mapping relationship mechanism of "data set-data record-information entity" is established, and the technical framework of agricultural science data ontology network system is designed. The system realizes the functions of automatic import, automatic management and ontology network visualization, breaks through the outstanding problems such as the large number of agricultural ontologies, the large amount of data, and the lack of proprietary systems for management, and develops an agricultural scientific data ontology network system that integrates four major functions, such as large-scale multi-format agricultural scientific data ontology import, ontology management, ontology and cross-ontology mapping relationship editing, and ontology network visualization, which effectively improves the management ability of agricultural scientific data ontology. It supports the efficient semantic association and release of massive data resources and the automatic aggregation of cross-domain and interdisciplinary data, which lays the foundation for online analysis and mining of agricultural scientific data.