Sentiment analysis based online restaurants fake reviews hype detection

Abstract

In our daily life, fake reviews to restaurants on e-commerce website have some great affects to the choice of consumers. By categorizing the set of fake reviews, we have found that fake reviews from hype make up the largest part, and this type of review always mislead consumers. This article analyzed all the characteristics of fake reviews of hype and find that the text of the review always tells us the truth. For the reason that hype review is always absolute positive or negative, we proposed an algorithm to detect online fake reviews of hype about restaurants based on sentiment analysis. In our experiment, reviews are considered in four dimensions: taste, environment, service and overall attitude. If the analysis result of the four dimensions is consistent, the review will be categorized as a hype review. Our experiment results have shown that the accuracy of our algorithm is about 74% and the method proposed in this article can also be applied to other areas, such as sentiment analysis of online opinion in emergency management of emergency cases. © Springer International Publishing Switzerland 2014.