Aspect based citation sentiment analysis using linguistic patterns for better comprehension of scientific knowledge

An almost unrestrained access to research plethora has emerged with a potential drawback: extracting relevant scientific publications is not a straightforward task anymore. The best way is to search on citation indexes, which also provide large number of pertinent papers and when a paper is focused even then it ascertains thousands of citations. In such a scenario, citation text could be a quintessential indicator in determining the importance and relevancy of paper for the researcher based on different aspects of the cited work such as technique, corpus, method, task, concept, measure, model and tool etc. This paper presents a novel approach to identify aspect level sentiments to reveal the hidden patterns from scholarly big data. The proposed methodology comprises of two levels. At first level, it extracts the aspects from the citation sentences using the pattern of opinionated phrases around the aspect. At the second level, it detects the sentiment polarity of the identified aspect considering nearby words and associates it with the corresponding aspect category based on a linguistic rule-based approach. We consider the words before, after and around the aspect using n-gram based features: ‘N-gram after’, ‘N-gram before’ and ‘N-gram around’. Our results reveal that ‘N-gram around’ feature performed better than other features and the SVM outperformed other considered classifiers for all N-gram models.