Sparsity and normalization in word similarity systems. (19th August 2015)

• Record Type: Journal Article
• Authors: GAWRON, JEAN MARK; STEPHENS, KELLEN
• Is Part Of: Natural language engineering. Volume 22:Part 3(2016)
• Page Start: 351
• Page End: 395
• Publication Date: 2015-08-19
• Subjects: Natural language processing (Computer science) -- Periodicals; Software engineering -- Periodicals; 006.35
• DOI: 10.1017/S1351324915000261
• Languages: English
• ISSNs: 1351-3249

Abstract

Abstract: We investigate the problem of improving performance in distributional word similarity systems trained on sparse data, focusing on a family of similarity functions we call Dice-family functions (Dice1945 Ecology 26 (3): 297–302), including the similarity function introduced in Lin (1998 Proceedings of the 15th International Conference on Machine Learning, 296–304), and Curran (2004 PhD thesis, University of Edinburgh. College of Science and Engineering. School of Informatics), as well as a generalized version of Dice Coefficient used in data mining applications (Strehl2000, 55). We propose a generalization of the Dice-family functions which uses a weight parameter α to make the similarity functions asymmetric. We show that this generalized family of functions (α systems) all belong to the class of asymmetric models first proposed in Tversky (1977 Psychological Review 84 : 327–352), and in a multi-task evaluation of ten word similarity systems, we show that α systems have the best performance across word ranks. In particular, we show that α-parameterization substantially improves the correlations of all Dice-family functions with human judgements on three words sets, including the Miller–Charles/Rubenstein Goodenough word set (Miller and Charles1991 Language and Cognitive Processes 6 (1): 1–28; Rubenstein and Goodenough1965 Communications of the ACM 8 : 627–633).