Construction of Deposit Model-oriented Knowledge Graph

It has become a hot research direction under the background of artificial intelligence to effectively organize geological big data and to mine corresponding potential value. The knowledge graph, as a new form of knowledge representation, can effectively organize and represent data. This paper presents the deposit model ontology according to the triune framework of “ ore-forming geological body, ore-forming structure and structural plane, and mineralization mark” by centering on the deposit model, sets up the knowledge graph with the “branch intersection, top-down action and step-by-step layering” method, expresses and mines big data for geological text in the form of knowledge graph, and establishes a special knowledge graph for deposit model with semantic processing capacity and open interconnection capacity, providing a routine to match geological data from new perspective and showing the integrity and relevance of genesis features of deposits on deeper level. The results show this method can effectively realize semantic correlation and constraints between deposit model ontology, as well as retrieval and reasoning of correlation objects, and prove the usefulness of the research workflow in this paper.