An Efficient method for Aspect Based Sentiment Analysis Using SpaCy and Vader

Many types of research have been carried out since today regarding Sentiment Analysis. But, very few amongst them have been able to move forward from sentiment analysis to a very useful concept of Aspect-Based Sentiment Analysis. Sentiment Analysis is typically done on three major levels, namely, Document-level, Sentence-level, and Aspect-level. The need for Aspect-Based Sentiment Analysis was raised when people understood that a whole text may have different types of sentiments related to different entities. Although, good results were derived from the Sentiment Analysis but digging more into the sentiment part was becoming difficult by following the same old methods. Aspects are the entities on which the reviews are given and different sentiments can be associated with different entities. Therefore, this research work is carried out to ease the things related to aspect-based sentiment analysis. In this research, an NLP (Natural Language Processing) model is trained on the text/reviews, and then the sentiment and aspects are mines using the vader, textblob, and spaCy library respectively. This research work will help highly in speeding-up the study of Aspect-Based Sentiment Analysis and will ease the struggle in mining the aspects in the text in accordance to the sentiment.