Review on Aspect based Sentiment Analysis on Social Data

Aspect-based sentiment analysis is a computationally determining opinions, sentiments, and semantic conception of the text. Alternatively, it is a specific sentiment analysis process with an aim to extract important features that describe the entities along with its polarity of every aspect present in the textual content. This paper reviews the recent research work in aspect based sentiment analysis to extract aspects in the text and corresponding sentiment. The paper discusses various aspect methods and sentiment prediction methods. The accuracy of IMAN is observed as 83.59 on Restaurant. The CNN-GA Hybrid model is observed with increased precision, accuracy, recall, and F1 calculation. Finally, we identify research opportunities in aspect or feature-based sentiment analysis.