Word Sense Selection in Texts An Integrated Model

Early systems for word sense disambiguation WSD often depended on individual tailor made lexical resources hand coded with as much lexical information as needed but of severely limited vocabulary size Recent studies tend to extract lexical informa tion from a variety of existing resources e g machine readable dictionaries corpora for broad coverage However this raises the issue of how to combine the information from di erent resources Thus while di erent types of resource could make di erent contribution to WSD studies to date have not shown what contribution they make how they should be combined and whether they are equally relevant to all words to be disambiguated This thesis proposes an Integrated Model as a framework to study the inter relatedness of three major parameters in WSD Lexical Resource Contextual Information and Nature of Target Words We argue that it is their interaction which shapes the e ectiveness of any WSD system A generalised structurally based sense mapping algorithm was designed to com bine various types of lexical resource This enables information from these resources to be used simultaneously and compatibly while respecting their distinctive structures In studying the e ect of context on WSD di erent semantic relations available from the combined resources were used and a recursive ltering algorithm was designed to overcome combinatorial explosion We then investigated from two directions how the target words themselves could a ect the usefulness of di erent types of knowledge In particular we modelled WSD with the cloze test format i e as texts with blanks and all senses for one speci c word as alternative choices for lling the blank A full scale combination of WordNet and Roget s Thesaurus was done linking more than senses Using these two resources in combination a range of dis ambiguation tests was done on more than noun instances from corpus texts of di erent types and blanks from real cloze texts Results show that combining resources is useful for enriching lexical information and hence making WSD more e ective though not completely Also di erent target words make di erent demand on contextual information and this interaction is closely related to text types Future work is suggested for expanding the analysis on target nature and making the com bination of disambiguation evidence sensitive to the requirements of the word being disambiguated