Processing on Structural Data Faultage in Data Fusion

Abstract

With the rapid development of information technology, the development of information management system leads to the generation of heterogeneous data. The process of data fusion will inevitably lead to such problems as missing data, data conflict, data inconsistency and so on. We provide a new perspective that combines the theory in geology to conclude such kind of data errors as structural data faultage. Structural data faultages after data integration often lead to inconsistent data resources and inaccurate data information. In order to solve such problems, this article starts from the attributes of data. We come up with a new solution to process structural data faultages based on attribute similarity. We use the relation of similarity to define three new operations: Attribute cementation, Attribute addition, and Isomorphous homonuclear. Isomorphous homonuclear uses digraph to combine attributes. These three operations are mainly used to handle multiple data errors caused by data faultages, so that the redundancy of data can be reduced, and the consistency of data after integration can be ensured. Finally, it can eliminate the structural data faultage in data fusion. The experiment uses the data of doctoral dissertation in Shanghai University. Three types of dissertation data tables are fused. In addition, the structural data faultages after fusion are processed by the new method proposed by us. Through the statistical analysis of the experiment results and compare with the existing algorithm, we verify the validity and accuracy of this method to process structural data faultages.