Ant Colony Algorithm for the Unsupervised Word Sense Disambiguation of Texts: Comparison and Evaluation
Didier Schwab, Jérôme Goulian, Andon Tchechmedjiev, Hervé Blanchon · 2012
Brute-force word sense disambiguation (WSD) algorithms based on semantic relatedness are really time consuming. We study how to perform WSD faster and better on the span of a text. Several stochastic algorithms can be used to perform Global WSD. We focus here on an Ant Colony Algorithm and compare it to two other methods (Genetic and Simulated Annealing Algorithms) in order to evaluate them on the Semeval 2007 Task 7. A comparison of the algorithms shows that the Ant Colony Algorithm is faster than the two others, and yields better results. Furthermore, the Ant Colony Algorithm coupled with a majority vote strategy reaches the level of the first sense baseline and among other systems evaluated on the same task rivals the lower performing supervised algorithms. TITLE AND ABSTRACT IN FRENCH Algorithme à colonie de fourmis pour la désambiguïsation lexicale non supervisée de textes: comparaison et évaluation Les algorithmes exhaustifs de désambiguïsation lexicale ont une complexité exponentielle et le contexte qu’il est calculatoirement possible d’utiliser s’en trouve réduit. Il ne s’agit donc pas