Towards automatic extraction of epistemic items from scientific publications

Abstract

The exponential growth of the World Wide Web in the last decade, brought with it an explosion in the information space. One, heavily affected area is the scientific literature, where finding relevant work in a particular field, and exploring links between relevant publications represents a cumbersome task. In this paper we make the initial steps in the direction of automatic extraction of epistemic items (i.e. claims, positions, arguments) from scientific publications. Our approach will provide the foundation for a comprehensive solution that will partly alleviate the information overload problem. We detail the actual extraction process, the evaluation we have performed and relevant use-cases for our work. © 2010 ACM.