DDL OMNI Engineering and Virginia Polytechnic Institute and State University will use our Salient Sentence Identification in Text (S2I2T) algorithm for the automated identification and characterization of the salient context related content in text documents. Our approach uniquely identifies salience from sentences within the documents, producing an easy to read result. Further, our approach allows the user to define the context directly at execution of the algorithm to match user entered terms against sentences identified as salient. During Phase I our approachs effectiveness will be validated via experiments comparing algorithmic results to human recognized salience. Additionally, we will continue refinements to our S2I2T prototype to improve algorithm performance and statistically validate the performance improvements.
Keywords: Information Value; Ontology; Semantic Web; Decision Support, Saliency, Data Fusion, Lexical Analysis
