Automated Aspect Extraction and Aspect Oriented Sentiment Analysis on Hotel Review Datasets

Abstract

Humans always try to make the best of their money be it food, clothes or hotel. Reviews provided by people play a crucial role in deciding if the hotel is worth staying. A hotel has many different aspects to be judged on while reviewing like food, rooms, service, etc. Due to the huge volume of available data on reviews it becomes a cumbersome task to summarize the gist of reviews. Hence, a need for an automated rating system arises. We have implemented an automated system that extracts the aspect terms from reviews and performs aspect-oriented sentiment analysis to summarize the review in a rating system on the basis of prominence of aspect terms.