This paper presents a novel unsupervised methodology for automatic disambiguation of nouns found in unrestricted corpora. The proposed method is based on extending the context of a target word by querying the web, and then measuring the overlap of the extended context with the topic signatures of the different senses by using Bayes rule. The algorithm is evaluated on Semcor 2.0. The evaluation showed that the web-based extension of the target word's local context increases the amount of contextual information to perform semantic interpretation, in effect producing a disambiguation methodology, which achieves a result comparable to the performance of the best system in SENSEVAL 3.