Conference paper

 
Otrusina, L., Smrž, P.: A New Approach to Pseudoword Generation, In: Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10), Valletta, MT, ELRA, 2010, p. 1-5, ISBN 2-9517408-6-7
Publication language:english
Original title:A New Approach to Pseudoword Generation
Title (cs):Nový přístup ke generování pseudo-slov
Pages:1-5
Proceedings:Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10)
Conference:Language Resources and Evaluation Journal 2010
Place:Valletta, MT
Year:2010
ISBN:2-9517408-6-7
Publisher:European Language Resources Association
Keywords
pseudo-words, word sense disambiguation, evaluation, SVM
Annotation
Sense-tagged corpora are used to evaluate word sense disambiguation (WSD)systems. Manual creation of such resources is often prohibitively expensive.That is why the concept of pseudowords - conflations of two or moreunambiguous words - has been integrated into WSD evaluation experiments. Thispaper presents a new method of pseudoword generation which takes into accountsemantic-relatedness of the candidate words forming parts of the pseudowords tothe particular senses of the word to be disambiguated. We compare the newapproach to its alternatives and show that the results on pseudowords, that aremore similar to real ambiguous words, better correspond to the actual results.Two techniques assessing the similarity are studied - the first one takesadvantage of manually created dictionaries (wordnets), the second one builds onthe automatically computed statistical data obtained from large corpora. Prosand cons of the two techniques are discussed and the results on a standard taskare demonstrated.
BibTeX:
@INPROCEEDINGS{
   author = {Lubomír Otrusina and Pavel Smrž},
   title = {A New Approach to Pseudoword Generation},
   pages = {1--5},
   booktitle = {Proceedings of the Seventh conference on International
	Language Resources and Evaluation (LREC'10)},
   year = {2010},
   location = {Valletta, MT},
   publisher = {European Language Resources Association},
   ISBN = {2-9517408-6-7},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=9193}
}