Human Fake review detection
Deception cues for detecting fake reviews in recommender systems
The continued digitization of business and private life also brings cybercriminals to the online world. One of the tactics of online fraudsters are fake consumer reviews on shopping platforms, such as Amazon, to either promote or demote a product. Research shows that most people who purchase a product or service online rely heavily on the opinion of others and base their decision on these potentially fake online reviews. Therefore, it is important that consumers can differentiate between real and fake reviews and come to informed decisions. However, there is very little research about how people determine whether the reviews they read are real or fake.
To close this gap, the PhD project of Michelle Walther aims to investigate the detection process of fake online reviews from the consumers perspective. The goal is to develop a new theoretical framework on how consumers detect and use (fake) reviews when shopping online and to then use the framework to develop meaningful interventions for consumers to help protect them from fake reviews.
We use a mixed methods approach, starting with a literature review, then interviews with consumers to build our initial theoretical framework. Next, we will triangulate the results with a thinking-aloud study and two experimental surveys to test and refine the developed theory.
Initial results show, that most people not only vary their purchasing behavior depending on the price and purpose of a product, but that they are also aware of fake reviews and customize their deception cues accordingly. The deception cues can be classified into categories including reviewer characteristics, content of the review, rating of the review, wording of the review, seller of the product and product characteristics. When comparing the results with earlier research on fake review detection one can see that though there are commonalities, there are also new detection cues found in the first user study.
If you are interested in this project please contact Michelle Walther or Steven Watson.