“ Recensione ” A Modified Rating Algorithm for Restaurants

Abstract

More and more people, now-a-days, make their purchase decision by referring to reviews and ratings provided in the online platform. It is evident that the visitors to restaurants use this facility in a larger scale compared with the users of other industries. However, evaluating the information so gathered, is a hazel and time consuming as it involves a process of reading through all the reviews, identifying the date of review posting, and understanding the reviewer’s credibility before making the decision. As a solution to this problem, the research proposes an enhanced rating algorithm which will take into account the following to calculate an overall rating which are aspect, sentiment, time factor and user credibility of the review. This enhanced algorithm uses Natural Language Processing and Sentiment analysis to identify the thoughts of the user regarding the restaurants. This will be a web-based solution that gives an overall idea of the current performance of a particular restaurant that will guide the users to choose a restaurant based on the overall rating with minimum hazel and time.