Deception detection and opinion spam

Abstract

In this chapter we first introduce the reader to the problem of deception detection in general, describing how lies may be detected automatically using different methods. Later we address the specific problem of deception detection in predatory communication. We make emphasis especially on those approaches using affective resources as categorical and psychometric information provided by natural language processing tools. Finally, we focus on the problem of opinion spam whose detection is very important for reliable opinion mining. In fact, nowadays a large number of opinion reviews are posted on theWeb. Such reviews are a very important source of information for customers and companies. Unfortunately, due to the business behind it, there is an increasing number of deceptive opinions on the Web. Those opinions are fictitious and have been deliberately written to sound authentic in order to deceive the consumers promoting a low quality product (positive deceptive opinions) or criticizing a potentially good quality one (negative deceptive opinions). Then, we summary some interesting approaches to detect spam opinion on the Web.