From owner-cse575-fa08-list@LISTSERV.BUFFALO.EDU Wed Oct 29 15:57:08 2008 Date: Wed, 29 Oct 2008 15:47:46 -0400 From: "William J. Rapaport" Subject: UB CogSci 11/5 Graeme Hirst: Semantic-Distance Measures Comments: To: COGSCI-ALL-LIST@LISTSERV.BUFFALO.EDU To: CSE575-FA08-LIST@LISTSERV.BUFFALO.EDU ======================================================================== NEXT WEEK NEXT WEEK NEXT WEEK NEXT WEEK NEXT WEEK NEXT WEEK ======================================================================== Center for Cognitive Science and Department of Computer Science & Engineering, University at Buffalo present ------------------------------------------------------------------------ Graeme Hirst Department of Computer Science University of Toronto http://www.cs.toronto.edu/~gh/ ------------------------------------------------------------------------ Wednesday, 5 November 2008; 2:00 p.m.; Park 280 ------------------------------------------------------------------------ Semantic-Distance Measures with Distributional Profiles of Coarse-Grained Concepts ------------------------------------------------------------------------ ABSTRACT: Although semantic-distance measures are applied to words in textual tasks such as building lexical chains, semantic distance is really a property of concepts, not words. We present a hybrid measure of semantic distance based on distributional profiles of concepts that we infer from text corpora. We use only a very coarse-grained inventory of concepts--each category of a published thesaurus is taken as a single concept--and yet obtain results on basic semantic-distance tasks that are generally as good as methods that use fine-grained, word-based measures. Because the measure is based on naturally occurring text, it is able to find word pairs that stand in non-classical relationships not found in WordNet. It can be applied cross-lingually, using a thesaurus in one language to measure semantic distance between words in another. In addition, it can used to determine the degree of antonymy between words. (Work done in association with Saif Mohammad, Institute for Advanced Computer Studies, University of Maryland.) BACKGROUND READING: Budanitsky, Alexander; & Hirst, Graeme (2006), "Evaluating WordNet-Based Measures of Lexical Semantic Relatedness", Computational Linguistics 32(1) (March): 13-47. http://www.mitpressjournals.org/doi/abs/10.1162/coli.2006.32.1.5 https://ublearns.buffalo.edu/@@3E63CDC03668CFCD339599F02EB3E84E/courses/1/ADM_PSY_MAUNER_081508/content/_1055641_1/Budanitsky%26Hirst2006CompLing.pdf ------------------------------------------------------------------------ Complete Fall schedule @ www.cogsci.buffalo.edu/Activities/Colloquium/CLLQf08/2008fall.html Background readings for each lecture are available to UB faculty and students on UB Learns. Once you have logged in to UB Learns at https://ublearns.buffalo.edu/ select "Center for Cognitive Science", then "Course Documents", then "Background Readings for Fall 2008 Colloquium Series". Or link directly to: https://ublearns.buffalo.edu/webapps/blackboard/content/listContent.jsp?course_id=_60597_1&content_id=_1010420_1 ------------------------------------------------------------------------ For further information, please visit: http://www.cogsci.buffalo.edu/ or contact: William J. Rapaport Webmaster, Center for Cognitive Science Associate Professor of Computer Science Affiliated Faculty, Philosophy & Linguistics 201 Bell Hall | (716) 645-3180 x 112 Dept. of Computer Science & Engineering | fax: (716) 645-3464 University at Buffalo (SUNY) | rapaport@cse.buffalo.edu Buffalo, NY 14260-2000 | http://www.cse.buffalo.edu/~rapaport Buffalo Restaurant Guide: http://www.cse.buffalo.edu/restaurant.guide/ Good Things about Buffalo: http://www.cse.buffalo.edu/~rapaport/buffalo.html