Mining Context Specific Similarity Relationships Using The World Wide Web

Proceedings of Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing (HLT/EMNLP), pages 499–506, Vancouver, October 2005. c⃝2005 Association for Computational Linguistics Mining Context Specific Similarity Relationships Using The World Wide Web Dmitri Roussinov Leon J. Zhao Weiguo Fan Department of Information Systems W.P. Carey School of Business Department of Management Information Systems Department of Information Systems Arizona State University University of Arizona Virginia Tech Tempe, AZ, 85287 Tucson, AZ 85721 Blacksburg, VA 24061 dmitri.roussinov@asu.edu lzhao@bpa.arizona.edu wfan@vt.edu Abstract We have studied how context specific web corpus can be automatically created and mined for discovering semantic similarity relationships between terms (words or phrases) from a given collection of documents (target collection). These relationships between terms can be used to adjust the standard vectors space representation so as to improve the accuracy of similarity computation between text documents in the target collection. Our experiments with a standard test collection (Reuters) have revealed the reduction of similarity errors by up to 50%, twice as much as the improvement by using other known techniques. 1 Introduction Many modern information management tasks such as document retrieval, clustering, filtering and summarization rely on algorithms that compute similarity between text documents. For example, clustering algorithms, by definition, place documents similar to each other into the same cluster. Topic detection algorithms attempt to detect documents or passages similar to those already presented to the users. “Query by example” retrieval is based on similarity between a document selected as example and the other ones in the collection. Even a classical retrieval task can be formulated as rank ordering according to the similarity between the document (typically very short) representing user’s query and all the documents in the collection. For similarity computation, text documents are represented by terms (words or phrases) that they have, and encoded by vectors according to a predominantly used vector space model (Salton & McGill, 1983). Each coordinate corresponds to a term (word or phrase) possibly present within a document. Within that model, a high similarity between a pair of documents can be only indicated by sharing same terms. This approach has apparent limitations due to the notorious vocabulary problem (Furnas et al., 1997): people very often use different words to describe semantically similar objects. For example, within a classical vector space model, the similarity algorithm would treat words car and automobile as entirely different, ignoring semantic similarity relationship between them. It has been known for a long time that semantic similarity relationships between terms can be discovered by their co-occurrence in the same documents or in the vicinity of each other within documents (von Rijsbergen, 1977). Until the 1990s, the studies exploring co-occurrence information for building a thesaurus and using it in automated query expansion (adding similar words to the user query) resulted in mixed results (Minker et al., 1972; Peat & Willett, 1991). The earlier difficulties may have resulted from the following reasons: 1) The test collections were small, sometimes only few dozens of documents. Thus, there was only a …

Content truncated to 500 words