Title:Document Retrieval using Predication Similarity

Abstract:Document retrieval has been an important research problem over many years in the information retrieval community. State-of-the-art techniques utilize various methods in matching documents to a given document including keywords, phrases, and annotations. In this paper, we propose a new approach for document retrieval that utilizes predications (subject-predicate-object triples) extracted from the documents. We represent documents as sets of predications. We measure the similarity between predications to compute the similarity between documents. Our approach utilizes the hierarchical information available in ontologies in computing concept-concept similarity, making the approach flexible. Predication-based document similarity is more precise and forms the basis for a semantically aware document retrieval system. We show that the approach is competitive with an existing state-of-the-art related document retrieval technique in the biomedical domain.