Sentiment Analysis of Yelp using Advanced V Model

Abstract

Yelp is a business review site that helps people find relevant places, based on the informed opinions of its users. It uses a five-point star rating system with user reviews. The main aim of the paper is to analyze sentiments on the reviews and devise a new five-scale sentiment rating to match it to the star rating that the user provided. Following the n-gram approach, this paper analyzes the adjective and adverbs occurring sequences and determines the trend or pattern of occurrence. After calculating a weighted score for each review, users’ ratings are predicted by treating multi class classification using different classifiers. The objective of the research in this paper is to try and determine how accurate the users are, in giving the ratings by comparing their star based on the calculated results.