Rule-Based Approach for Party-Based Sentiment Analysis in Legal Opinion Texts 
2020 20th International Conference on Advances in ICT for Emerging Regions (ICTer)
Aspect Based Sentiment Analysis (ABSA) deals with extracting aspects from a given text and then allocate each aspect a sentiment level (positive, negative or neutral) [1]. A number of researchers have addressed ABSA in different domains except for the legal domain. In this study, we explore how the concepts related to aspect-based sentiment analysis can be used in the legal domain to extract valuable information from legal opinion text. To this regard, we propose a rule based approach to perform aspect-based sentiment analysis in order to figure out the sentiment level of a sentence in relation to each legal party related to a court case (considering the legal parties as the aspects).
Keywords: Natural Language Processing | Machine Learning / Deep Learning | Law | Legal Domain | Sentiment Analysis |