A Semantic Similarity Approach to Electronic Document Modeling and Integration

The World Wide Web is an enormous collection of information resources serving for various purposes. However the diversity of the Web information as well as the relared formats makes it vety difficult for users to efficiently search and obtain the information they require, The reason for the difficulty is because most of the information uploaded to the Web is unstructured or less structured. Many metadata models are proposed to response to this problem. These metadata models attempt to provide a certain kind of general description for the Web infimnation to improve its structuredness. Although these documents consist in a largest portion of the Web informarion or Web resources, few metadata models are dealing with the ill-structured Web documents through analyzing their semantic relations with each other. In this paper we consider this large portion of the Web information, called electronic documents. We propose a metadara model, called EDM (Electronic Document Metadata Model). Using the metadata model we can extract semantic characteristics from electronic documents and then use the characteristics to form a semantic electronic document model. This model, inverseiy, provides a basis for analysis of semantic similarity between electronic documents and for the electronic document integration. The document modeling and integration will support further manipulations on the electronic documents, such as exchange, search, and evolution.