Bio2X: a rule-based approach for semi-automatic transformation of semi-structured biological data to XML

摘要

Data integration of geographically dispersed, heterogeneous, complex biological databases is a key research area. One of the key features of a successful data integration system is to have a simple self-describing data exchange format. However, many of the biological databases provide data in flat files which are poor data exchange formats. Fortunately, XML can be viewed as a powerful data model and better data exchange format. In this paper, we present the Bio2X system that transforms flat file data into highly hierarchical XML data using rule-based machine learning technique. Bio2X has been fully implemented using Java. Our experiments to transform real world biological data demonstrate the effectiveness of the Bio2X approach.