Sentiment Analysis for Hotel Reviews

Home / Artificial Intelligence & ML / Sentiment Analysis for Hotel Reviews

Sentiment Analysis for Hotel Reviews

Travel planning and hotel booking on the website have become one of an important commercial use. Sharing on the web has become a major tool in expressing customer thoughts about a particular product or Service. Recent years have seen rapid growth in online discussion groups and review sites (e.g.www.tripadvisor.com) where a crucial characteristic of a customer’s review is their sentiment or overall opinion — for example, if the review contains words like ‘great’, ‘best’, ‘nice’, ‘good’, ‘awesome’ is probably a positive comment. Whereas if reviews contain words like ‘bad’, ‘poor’, ‘awful’, ‘worse’ is probably a negative review. However, Trip Advisor’s star rating does not express the exact experience of the customer. Most of the ratings are meaningless, a large chunk of reviews fall in the range of 3.5 to 4.5 and very few reviews below or above. We seek to turn words and reviews into quantitative measurements. We extend this model with a supervised sentiment component that is capable of classifying a review as positive or negative with accuracy (Section 4). We also determine the polarity of the review that evaluates the review as recommended or not recommended using semantic orientation. A phrase has a positive semantic orientation when it has good associations (e.g., “excellent, awesome”) and a negative semantic orientation when it has bad associations (e.g., “terrific, bad”). Next step is to assign the given review to a class, positive or negative, based on the average semantic orientation of the phrases extracted from the review. If the average is positive, the prediction is that the review posted is positive. Otherwise, the prediction is that the item is negative.

Research Paper Link: Download Paper

Related Post

Leave a Reply