A Trustworthiness Scoring Model for Reviews Using a Graph Structure

It refers to determine whether a product is good or bad at first glance because reviews mixedly contain positive or negative opinions. In this study, we propose a method to estimate users’ and stores’ trustworthiness using the network structure. Our method is based on the idea of eigenvector centrality and the Hypertext Induced Topic Selection (HITS) algorithm. The average value of the user’s usefulness obtained from P Yelp dataset tended to be higher when the user’s trustworthiness was higher.