Social Data Analysis: A Study on Friend Rating Influence

Social Networking accounts for a significant chunk of interest among various online activities [11]. The proclivity of being social, online, has been ingrained in us so much that we are actively producing content for the rest of the world to see or take interest in our whereabouts, our meals, our opinions, photographs etc. Yelp, seamlessly, integrates this very aspect of people in its portal. It engages people to write reviews about the businesses they have availed the services of, rate them, add photographs, tags, follow other people and their activities, etc. In this paper we examine and present the co-relation between a user’s rating and the influence of the people, that the user follows, on the user for a particular business. The group of people that the user follows is commonly referred as friends of the user. We also analyze if a user can get influenced, if a business has a certain number of reviews already present or if the reviews have been written by elite reviewers (a reviewer who, according to Yelp, has contributed exceptionally in engaging the community in the form of consistency in writing reviews, as well as the quality of the reviews). Our analysis, through correlation and regression techniques, is able to prove that the user’s rating remains unaffected by the number of people a user was friends with nor does the existing number of reviews and presence of elite reviewers helps in influencing a user. What shapes a user’s rating is the overall experience, that the user had at the restaurant. Keywords—Yelp data analysis, social network analysis, regression, visualization, correlation