LREC 2000 2nd International Conference on Language Resources & Evaluation
 

Previous Paper   Next Paper

Title An Approach to Lexical Development for Inflectional Languages
Authors Turcato Davide (Natural Language Laboratory, School of Computing Science, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, V5A 1S6, Canada, fturk@cs.sfu.ca)
Toole Janine (Natural Language Laboratory, School of Computing Science, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, V5A 1S6, Canada, toole@cs.sfu.ca)
Tsiplakou Stavroula (Department of Linguistics, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, V5A 1S6, Canada, fstavroula_tsiplakou@sfu.ca)
Heift Trude (Department of Linguistics, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, V5A 1S6, Canada, heiftg@sfu.ca)
McFetridge Paul (Natural Language Laboratory, School of Computing Science, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, V5A 1S6, Canada, mcfetg@cs.sfu.ca)
Keywords Computational Lexicons, Corpus-Based Techniques, Inflectional Languages, Morphology
Session Session WO18 - Morphology in Lexical and Textual Resources
Full Paper 256.ps, 256.pdf
Abstract We describe a method for the semi-automatic development of morphological lexicons. The method aims at using minimal pre-existing resources and only relies upon the existence of a raw text corpus and a database of inflectional classes. No lexicon or list of base forms is assumed. The method is based on a contrastive approach, which generates hypothetical entries based on evidence drawn form a corpus, and selects the best candidates by heuristically comparing the candidate entries. The reliance upon inflectional information and the use of minimal resources make this approach particularly suitable for highly inflectional, lower-density languages. A prototype tool has been developed for Modern Greek.