Ontology-driven relational data mapping for constructing a knowledge graph of porphyry copper deposits

Abstract

Geoscience knowledge graph has become a popular topic in recent years. A series of studies have been reported to introduce the construction and application of geoscience knowledge graphs from different views. The relational geoscience dataset with high knowledge density and data quality is an important digital heritage of geoscience. The relational dataset has not been taken seriously in the geoscience knowledge graph research. In this study, we proposed a quick method of building a geoscience knowledge graph using relational data mapping to triples. First, the use-case-driven method was applied to design the ontology of porphyry copper deposits. Second, the mapping rules were built based on the porphyry copper ontology. Third, the knowledge graph of the porphyry copper deposit was constructed based on relational data mapping and knowledge fusion. Based on the resulting knowledge graph, several exploratory cases were conducted to make knowledge reasoning and discovery. It is indicated that the solution proposed in this study is a fast batch-processing geoscience knowledge graph construction method. The experiences from this study can benefit the construction of knowledge graphs in other geoscience disciplines and promote knowledge discovery.