AI-Based Emotion Recognition to Study Users’ Perception of Dark Patterns

Dark Patterns are design patterns used to trick users into acting against their real interest. The web provides an infinite number of services accessible to anyone, which do not always promote a good user experience and are often structured with the aim of leading the user to perform unwanted actions or discourage him from making decisions that could damage the company. This is a very common practice, especially in neuromarketing. Human behavioral and perceptual patterns are cleverly exploited to achieve a specific goal. For this reason, dark pattern developers try to create an environment that invites as much purchase as possible by stimulating the customer’s unconscious. Among the areas in which these strategies are adopted is tourism: online travel agency websites promote “fake discounts” for the products/services they are selling, display inaccurate pricing information leading to incorrect pricing assumptions, thus misleading consumers. One of the goals of this work is to identify which dark patterns are most used in online travel agencies and once identified, they will be used to run scenarios that will simulate booking a vacation online. During the execution of the tests, users will be filmed via webcam tracking their expressions and emotions through AI-based facial recognition. Finally, the data obtained from the tests will be analyzed to study the emotions and feelings that a user feels when he/she is confronted with dark patterns, to understand which users are more at risk and which are the types of dark patterns to which they are more vulnerable.