On entropy-based term weighting schemes for text categorization

Abstract

Term weighting schemes often dominate the performance of many classifiers, such as kNN, centroid-based classifier and SVMs. The widely used term weighting scheme in text categorization, i.e., tf.idf, is originated from information retrieval (IR) field. The intuition behind idf for text categorization seems less reasonable than IR. In this paper, we introduce inverse category frequency (icf) into term weighting scheme and propose two novel approaches, i.e., tf.icf and icf-based supervised term weighting schemes. The tf.icf adopts icf to substitute idf factor and favors terms occurring in fewer categories, rather than fewer documents. And the icf-based approach combines icf and relevance frequency (rf) to weight terms in a supervised way. Our cross-classifier and cross-corpus experiments have shown that our proposed approaches are superior or comparable to six supervised term weighting schemes and three traditional schemes in terms of macro-F1 and micro-F1.

A new simple and effective measure for bag-of-word inter-document similarity measurement

“…In the supervised text mining task of document classification, different approaches utilizing class information are proposed to estimate the collection-based term weighting factors (Wang and Zhang, 2013;Debole and Sebastiani, 2003;Lan et al, 2009). Inverse category frequency (icf) (Wang and Zhang, 2013) has been shown to produce better classification result than the traditional idf factor with the cosine similarity measure. It considers the distribution of a term among classes rather than among documents in the given collection.…”

“…It considers the distribution of a term among classes rather than among documents in the given collection. The intuition behind icf is that the fewer classes a term t i occurs in, the more discriminating power the term t i contributes to classification (Wang and Zhang, 2013). If C and c i are the total number of classes and the number of classes in which t i occurs at least once in at least one document, then the icf factor is estimated as: icf (t i ) = log 1 + C c i .…”

“…If C and c i are the total number of classes and the number of classes in which t i occurs at least once in at least one document, then the icf factor is estimated as: icf (t i ) = log 1 + C c i . We have evaluated the performance of Sp in the kNN document classification task with existing measures using the supervised term weighting scheme of icf (Wang and Zhang, 2013). Sp produced either better or competitive classification results with existing measures using supervised or unsupervised term weighting in the 5NN classification task.…”

“…In the field of IR, there has been considerable research investigating the effective term weighting scheme. The importance of a term t i in document x, w i (x), is estimated using different variants and combinations of two factors (Manning et al, 2008;Salton and Buckley, 1988;Joachims, 1997;Robertson et al, 1994;Singhal, 1997;Roberston and Zaragoza, 2009;Paltoglou and Thelwall, 2010;Han et al, 2012;Wang and Zhang, 2013): (i) documentbased factor based on the frequency of t i in x, x i ; and (ii) collection-based factor based on the number of documents where t i occurs, n i . The most widely used term weighting scheme is tf-idf (term frequency -inverse document frequency) where w i (x) = tf i (x)× idf (t i ) (Manning et al, 2008;Salton and Buckley, 1988); and it consists of: i. Document-based factor: tf i (x) = 1 + log(x i ) if x i > 0, and 0 otherwise; ii.…”

“…Prior research in the BoW inter-document similarity measurement task were focused on developing effective term weighting schemes to improve the task specific performances of existing measures such as cosine and Best Match 25 (BM25) (Salton and Buckley, 1988;Robertson et al, 1994;Joachims, 1997;Singhal, 1997;Roberston and Zaragoza, 2009;Paltoglou and Thelwall, 2010;Han et al, 2012;Wang and Zhang, 2013). In contrast, we investigate an alternative similarity measure where an adjustment of vector components using term weighting is not required.…”