Conditional Random Field Based Sentence Context Identification: Enhancing Citation Services for the Research Community

Angrosh M.A., Cranefield, S. and Stanger, N.

Academic publishers' full text databases are an important part of the deep Web for researchers and a potentially valuable resource for automated extraction of scientific knowledge. Recently, some major publishers have provided Web APIs for accessing their article databases, thus allowing the development of Web applications to mine these resources. However the task of knowledge discovery from academic articles, particularly with citations remains a challenge. We present in this paper our research work taken up for identifying contexts associated with sentences in academic articles and use of this information to provide information services for the research community. To this end, we propose an annotation scheme for sentences in academic articles. We also describe our experiments with conditional random fields for sentence classification. Finally, we present CitContExt � a citation context extraction application developed based on the techniques discussed above.

Cite as: Angrosh M.A., Cranefield, S. and Stanger, N. (2013). Conditional Random Field Based Sentence Context Identification: Enhancing Citation Services for the Research Community. In Proc. The Web 2013 (AWC 2013) Adelaide, Australia. CRPIT, 144. Ashman, H., Sheng, Q.Z. and Trotman, A. Eds., ACS. 59-68

(from crpit.com) (local if available)