Automated word sense disambiguation for web information retrieval
Christopher M. Stokoe · ACM SIGIR Forum · 2005
A word in the English language is considered ambiguous if, regardless of context, it can have more than one possible interpretation or meaning. Many words exhibit lexical ambiguity suggesting that it has the potential to impact upon the performance of text retrieval systems. This may be particularly true in the case of web retrieval given the hypothesis that short queries may not provide sufficient context to adequately differentiate between opposing meanings of constituent words. Word sense disambiguation is an active field of study which seeks to create software which automatically resolves ambiguity through mapping word use to meaning. In this study the author examined the use of word sense disambiguation in order to resolve ambiguity within an IR collection. The motivation behind this work was to demonstrate the potential for increased retrieval effectiveness as a result of performing word sense disambiguation.