Similarity Model and Term Association for Document Categorization

Abstract

In the information retrieval and document categorization context, both Euclidean distance- and cosine-based similarity models are based on the assumption that term vectors are orthogonal. But this assumption is not true. Term associations are ignored in such similarity models. This paper analyzes the properties of term-document space, term-category space and categorydocument space. Then, without the assumption of term independence, we propose a new mathematical model to estimate the association between terms and define a ∞-similarity model of documents. Here we make best use of existing category membership represented by corpus as much as possible, and the objective is to improve categorization performance. The empirical results been obtained by k-NN classifier over Reuters-21578 corpus show that utilization of term association can improve the effectiveness of categorization system and ∞- similarity model outperforms than ones without term association.