Knowledge-based Wrapper Generation by Using XML

Information extraction is the process of recognizing the particular fragments of a document that constitute its core semantic content. However, most previous information extraction systems were not effective for real-world information sources due to difficulties in acquiring and representing useful domain knowledge and in dealing with structural heterogeneity inherent in different sources. In order to resolve these problems, this paper proposes a scheme of knowledge-based wrapper generation for semi-structured and labeled documents. The implementation of an agent-oriented information extraction system, XTROS, is described. In contrast with previous wrapper learning agents, XTROS represents both the domain knowledge and the wrappers by XML documents to increase modularity, flexibility, and interoperability among multiple parties. XTROS also facilitates simpler implementation of the wrapper generator by exploiting XML parsers and interpreters. XTROS shows good performance on several Web sites in the domain of real estates, and it is expected to be easily adaptable to different domains by plugging in appropriate XML-based domain knowledge.