A relation mining and visualization framework for automated text summarization

Abstract

In this paper, we present a relation mining and visualization framework to identify important semi-structured information components using semantic and linguistic analysis of text documents. The novelty of the paper lies in identifying key snippet from text to validate the interaction between a pair of entities. The extracted information components are exploited to generate semantic network which provides distinct user perspectives and allows navigation over documents with similar information components. The efficacy of the proposed framework is established through experiments carried out on biomedical text documents extracted through PubMed search engine. © 2009 Springer-Verlag Berlin Heidelberg.