A Concept-based Analysis of Dark Patterns in User Interface Design

Abstract

In this thesis, we present a new theory of dark patterns that aims to move the literature away from fragmented, top-down analysis of systems through subjective dark patterns taxonomies towards a more objective and generalizable framework. Focusing on user interface related dark patterns, we propose a single similarity that they all have: the user interface misrepresents the underlying conceptual functionality of the system in a way that benefits the system owner at the cost of the user. We then present a domain- and modality-independent framework that breaks systems down into isolated, reusable units of functionality called concepts and systematically codifies user expectations for each concept and how it should be mapped to the user interface. We illustrate our framework on three popular concepts in e-commerce applications that a significant number of dark patterns stem from: Catalog, ShoppingCart, and MailingList. We show that the framework design allows us to apply our dark pattern definition in an objective manner, while capturing dark patterns previously identified in the literature related to these concepts and being extensible to analyzing new designs in the future. We conclude with possible use cases of the framework for researchers, industry designers, and legislators.