Good Neighbors Make Good Senses: Exploiting Distributional Similarity for Unsupervised WSD

We present an automatic method for sense-labeling of text in an unsupervised manner. The method makes use of distributionally similar words to derive an automatically labeled training set, which is then used to train a standard supervised classifier for distinguishing word senses. Experimental results on the Senseval-2 and Senseval-3 datasets show that our approach yields significant improvements over state-of-the-art unsupervised methods, and is competitive with supervised ones, while eliminating the annotation cost.