The Community for Technology Leaders
Green Image
Issue No. 06 - June (2013 vol. 25)
ISSN: 1041-4347
pp: 1307-1322
Yelena Yesha , University of Maryland, Baltimore County, Baltimore
Anupam Joshi , University of Maryland, Baltimore County, Baltimore
Tim Finin , University of Maryland, Baltimore County, Baltimore
Paul McNamee , Johns Hopkins University, Baltimore
Lushan Han , University of Maryland, Baltimore County, Baltimore
Pointwise mutual information (PMI) is a widely used word similarity measure, but it lacks a clear explanation of how it works. We explore how PMI differs from distributional similarity, and we introduce a novel metric, $({\rm PMI}_{max})$, that augments PMI with information about a word's number of senses. The coefficients of $({\rm PMI}_{max})$ are determined empirically by maximizing a utility function based on the performance of automatic thesaurus generation. We show that it outperforms traditional PMI in the application of automatic thesaurus generation and in two word similarity benchmark tasks: human similarity ratings and TOEFL synonym questions. $({\rm PMI}_{max})$ achieves a correlation coefficient comparable to the best knowledge-based approaches on the Miller-Charles similarity rating data set.
Context, Thesauri, Correlation, Semantics, Measurement, Vectors, Mathematical model, corpus statistics, Semantic similarity, pointwise mutual information, automatic thesaurus generation
Yelena Yesha, Anupam Joshi, Tim Finin, Paul McNamee, Lushan Han, "Improving Word Similarity by Augmenting PMI with Estimates of Word Polysemy", IEEE Transactions on Knowledge & Data Engineering, vol. 25, no. , pp. 1307-1322, June 2013, doi:10.1109/TKDE.2012.30
196 ms
(Ver 3.3 (11022016))