Building XML Data Warehouse with Data Reconstruction by Knowledge Graph

The analysis and decision scenarios, which are more complex in many domains, request more unstructured data to be reorganized with the structured data for analyzing. An architecture is proposed in this paper for data reconstruction to support OLAP analysis based on the knowledge graph and XML data warehouse technologies. In this architecture, the data cubes can be built with the data of two kinds of knowledge graphs, and the data populating of the cubes is on the basis of the correlation of the entities in knowledge graphs. Under the consideration of utilizing both information of the ontology concepts and their entities, a modified module in the architecture is also proposed for good concept selection to extract the cube dimension hierarchies. The evolution results prove the effectiveness and advantages of our proposed architecture and modified module.