Comparison of Algorithmic and Human Assessments of Sentence Similarity

This paper describes a new method, based on information theory, for measuring sentence similarity. The method first computes the information content (IC) of dependency triples using corpus statistics generated by processing the Open American National Corpus (OANC) with the Stanford Parser. We define the similarity of two sentences as a function of (1) the similarity of their constituent dependency triples, and (2) the position of the triples in their respective dependency trees. We apply the algorithm to 15 pairs of sentences that were also given to human subjects to assign a similarity score. The humanand computer-generated scores are compared; the results are promising, but point to the need for further refinement.