With the commencement of new technology and increase of online shopping companies like Amazon, E-Bay, and Flipkart, people give a wide range of reviews for the products they purchase. Some are too long, some too short, some difficult to understand while some are totally irrelevant. Thus, there is a pressing need to reduce the diversity of reviews for a particular product and to show the users only the gist of most useful and important reviews about a product. The proposed work focuses to summarize the reviews for movies purchased from Amazon using a combination of four state-of-the-arts algorithms and a feature selection technique. Sentiment analysis has been performed to categorize the reviews into positives and negatives. Also, a novel method named hierarchical summarization is proposed to summarize large reviews into summary of few sentences. The results of this summary are compared with the existing algorithms using the ROUGE score to determine the best summary. Experimental results show that the proposed approach is promising.
CITATION STYLE
Roul, R. K., & Sahoo, J. K. (2020). Sentiment Analysis and Extractive Summarization Based Recommendation System. In Advances in Intelligent Systems and Computing (Vol. 990, pp. 473–487). Springer. https://doi.org/10.1007/978-981-13-8676-3_41
Mendeley helps you to discover research relevant for your work.