In the last years, the amount of textual electronic information available has been increasing rapidly. Computer understanding of texts has become an important trend in computational linguistics. Proper processing of this kind of information requires an interpretation of their meaning at a semantic level. This work presents novel measures to estimate the degree of semantic similarity between words using one or more knowledge sources. The measures are based on the exploitation of the knowledge modelled in one or several ontologies and on the estimation of the information distribution of terms in the Web. They have been applied to clustering, computing the similarity/distance between individuals described by textual attributes. Results show that a proper interpretation of textual data at a semantic level improves the quality of the clusters and eases their interpretability.