1 Introduction
In natural language processing (NLP), word sense disambiguation (WSD) is the task of automatically determining the right sense of a polysemous word given its context [1], [2]. The problem of WSD is of great interest due to the intrinsic ambiguity of natural languages. Other fields of NLP can potentially benefit from advances in WSD, such as question answering [3], [4] or sentiment analysis [5], [6], just to mention two examples.