Word-sense disambiguation

Resnik, R, and D. Yarowsky. 2000. Distinguishing systems and distinguishing senses New evaluation methods for word sense disambiguation. Nat Lang Engineering 5 113-33. 

Yarowsky, D. 2000. Hierarchical decision lists for word sense disambiguation. Computers and the Humanities 34 179-86. 

The second type of algorithm performs word sense disambiguation (WSD). The difference between this and tagging is that in word sense disambiguation, we define multiple sets of labels Li,L2,...,Lm and choose a label from one of these sets for only some of the tokens. In natural language processing, WSD is often used for semantic disambiguation to resolve cases such as ... 

The naive Bayes classifier is a simple and popular approach in word sense disambiguation. In this, we attempt to estimate the probability that we have word w, is llie correct one for a token based on a number of features ... 

A basic distinction can be made between taggers which assign a label to every token drawn from a single set, and word sense disambiguators (WSD) which assign labels only to ambiguous tokens and do so from multiple label sets. 

Gale, W. A., Church, K. W., and Yarowsky, D. Using bilingual materials to develop word sense disambiguation methods. In International Conference on Theoretical and Methodological Issues in Machine Translation (1992), pp. 101-112. 

---