This seminar will present a machine learning approach for determining the role that individual citations play in scientific articles, so that this information can be used to improve search and browsing capabilities in scientific article databases. We begin with describing some of the current citation related tools and services and discuss some of the limitations of these services, particularly with reference to contextual information services for citations. We then present our tool as a possible solution for providing additional information about citations to users and talk about the key components of our tool, which includes an annotation scheme designed for sentences in scientific articles and the machine learning technique employed for achieving sentence classification.
Angrosh has recently completed a PhD in Information Science at the University of Otago. His research focused on identifying contexts of sentences in scientific articles and used the contextual information to develop value-added tools for the scientific community. Before starting his PhD, he was involved with the Vidyanidhi Digital Library (an e-thesis portal) at the University of Mysore, India and was also working on developing Semantic Web applications for the agri-pest domain.
Last modified: Tuesday, 26-Feb-2013 10:33:56 NZDT
This page is maintained by the seminar list administrator.