A New Approach to Pseudoword Generation

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.