Distributional Word Clusters vs. Words for Text Categorization

By Ron Bekkerman, Ran El-Yaniv, Naftali Tishby, Yoad Winter. Journal of machine learning research, 2003

348 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/a5000120bac6a3fabb875ce5d05af400d699b571

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Analysis of the IJCNN 2007 agnostic learning vs. prior knowledge challenge (2008, methodology): “To our knowledge, the best results on the 20 Newsgroup data were obtained by the method of distributional clustering by Ron Bekkerman [4].”
An introduction to variable and feature selection (2003, methodology): “A more elaborate approach is taken by Bekkerman et al. (2003) who use a “soft” version of K-means (allowing words to belong to several clusters) and who progressively divide clusters by varying the Lagrange multiplier β monitoring the tradeoff between I(X, X̃) and I(X̃, Y ).”
Modeling bug report quality (2007, methodology): “Recent work in automatic bug report analysis has tended to use sophisticated machine learning techniques, such as support vector machines [3] and clustering [17], in combination with text categorization techniques [4].”
Feature selection via regularized trees (2012, background): “nd Background 2.1 Related work Feature selection methods can be divided into ﬁlters, wrappers and embedded methods [4]. Filters select features based on criteria independent of any supervised learner [5, 6, 7, 8, 9, 10, 11, 12, 13]. Therefore, the performance of ﬁlters may not be optimum for a chose”
A survey on feature selection methods (2014, methodology): “In [13] twelve feature selection metrics are considered for the text classification problem [1,13,18].”
An Introduction to Variable and Feature Selection (2003, unknown): “Several approaches to the variable selection problem using information theoretic criteria have been proposed (as reviewed in this issue by Bekkerman et al., 2003, Dhillon et al., 2003, Forman, 2003, Torkkola, 2003).”
Sparse evolutionary deep learning with over one million artificial neurons on commodity hardware (2019, methodology): “Traditional feature selection methods can be roughly divided into three categories: filter methods (Guyon & Elisseeff, 2003; Forman, 2003; Bekkerman et al., 2003; Caruana & Sa, 2003; Koller & Sahami, 1996; Davidson & Jalan, 2010; de Jesus Rubio et al., 2019; de Jesús Rubio, 2009), wrapper methods (K”