Measures of semantic similarity have been traditionally defined between words or concepts, and much less between text segments consisting of two or more words. The emphasis on word-to-word similarity metrics is probably due to the vailability of resources that specifically encode relations between words or concepts (e.g. WordNet), and the various testbeds that llow for their evaluation (e.g. TOEFL or SAT analogy/synonymy tests). Moreover, the derivation of a text-to-text measure of imilarity starting with a wordbased semantic similarity metric may not be straightforward,and consequently most of the work in this area has considered mainly applications of the traditional vectorial model,occasionally extended to n-gram language models.