Semantic Based Clustering of Web Documents

A new methodology that structures the semantics of a collection of documents into the geometry of a simplicial complex is developed. A simplicial complex is topologically equivalent to a polyhedron in Euclidean space. The semantics of documents are structured by the geometry: A primitive concept is represented by a simplex. and a concept is represented by a connected component. Based on these structures, documents can be clustered into some meaningful classes. Experiments with three different data sets from web pages and medical literature have shown that our approach performs significantly better than traditional clustering algorithms, such as k-means, AutoClass and Hierarchical Clustering (HAC). keyword clustering, association(rule)s, topology, simplicial complex, polyhedron