A VSM-based Statistical Model for the Semantic Relation Interpretation of Noun-Modifier Pairs

The paper addresses the task of automatic interpretation of semantic relation in noun compounds. The problem has been attempted with both Ontology-based and Statistical approaches, but both approaches having their own limitations. We present a novel VSMbased statistical model which represents each relation with a weighted vector of prepositional and verbal paraphrases. The model ranks the paraphrases on their relevance and assigns higher weights to more relevant paraphrases. The performance of the model is compared with the Ontology model and the results are quite encouraging. We finally propose a Hybrid of the two models which compares on par with the best performing systems on Nastase and Szpakowicz (2003) dataset.