Abstract:
|
In this paper we apply Schapire and Singer's AdaBoost.MH boosting algorithm to the Word Sense Disambiguation (WSD)
problem. Initial experiments on a set of 15 selected polysemous words show that the boosting approach surpasses Naive Bayes
and Exemplar--based approaches, which represent state--of--the--art accuracy on WSD. In order to make boosting practical for a
real learning domain of thousands of words we study several ways of accelerating the algorithm by reducing the feature space.
The best variant, which we call LazyBoosting, is tested on a medium--large sense--tagged corpus containing 192,800 examples of
the 191 most frequent and ambiguous English words. Again, boosting compares favourably to the other benchmank algorithms. |