Selectional Preferenced Based Verb Sense Disambiguation Using WordNet

Abstract

Selectional preferences are a source of linguistic information commonly applied to the task of Word Sense Disambiguation (WSD). To date, WSD systems using selectional preferences as the main disambiguation mechanism have achieved limited success. One possible reason for this limitation is the limited number of semantic roles used in the construction of selectional preferences. This study investigates whether better performance can be achieved using the current state-of-art semantic role labelling systems, and explores alternative ways of applying selectional preferences for WSD. In this study, WordNet noun synonym sets and hypernym sets were used in the construction of selectional preferences; Semcor2.0 data was used for the training and evaluation of a support vector machine classifier and a Naive Bayes classifier.