Automatic classification of citation function

TLDR

Citation function refers to the author’s reason for citing a paper, and automatically recognizing it has applications from impact‑factor calculation to text summarisation. The study aims to demonstrate the reliability of an annotation scheme for citation function and to present a supervised machine‑learning framework for its automatic classification. The framework employs supervised learning with shallow and linguistically‑inspired features. The results reveal a strong relationship between citation function and sentiment classification.

Abstract

Citation function is defined as the author's reason for citing a given paper (e.g. acknowledgement of the use of the cited method). The automatic recognition of the rhetorical function of citations in scientific text has many applications, from improvement of impact factor calculations to text summarisation and more informative citation indexers. We show that our annotation scheme for citation function is reliable, and present a supervised machine learning framework to automatically classify citation function, using both shallow and linguistically-inspired features. We find, amongst other things, a strong relationship between citation function and sentiment classification.