A Model Driven Approach to Constructing Knowledge Graph from Relational Database

Shanxin Sun,Fanchao Meng,Dianhui Chu
DOI: https://doi.org/10.1088/1742-6596/1584/1/012073
2020-07-01
Journal of Physics: Conference Series
Abstract:Abstract Relational database is one of the main data sources in the construction of vertical domain knowledge graph. How to establish the mapping relationship between the concepts in relational database and the corresponding concepts in knowledge graph is the key to improve the efficiency of the construction of vertical domain knowledge graph. To solve this problem, this paper proposes a model driven approach to automatically constructing knowledge graph from relational database based on the Model Driven Architecture (MDA). First of all, a model driven framework is proposed to describe the relational database and knowledge graph using standard XML Schema. Then, a model driven approach is proposed, which extracts database models and data to XML, defines the conversion functions, transforms models and data from database to knowledge graph, and builds knowledge graph according to the expression forms. Finally, using a case to validate the approach. This paper uses the idea of model driven to realize the high-quality transformation from relational database to knowledge graph.
What problem does this paper attempt to address?