Detection of Implicit Citations for Sentiment Detection

Sentiment analysis of citations in scientific papers is a new and interesting problem which can open up many exciting new applications in bibliometrics. Current research assumes that using just the citation sentence is enough for detecting sentiment. In this paper, we show that this approach misses much of the existing sentiment. We present a new corpus in which all mentions of a cited paper have been annotated. We explore methods to automatically identify these mentions and show that the inclusion of implicit citations in citation sentiment analysis improves the quality of the overall sentiment assignment.