Investigating the effect of Transparency and Data Control on Trust, Privacy Concerns and the Perceived Quality of Deep Movie Recommender Systems

Abstrakt

Consumers are presented with a sea of different products and services, and the consumer decision-making process becomes overwhelming as a result. Recommender systems have been invented to solve this issue, alleviating the consumer from the stress that is navigating in and filtering the endless amount of information. Personalized recommendations have proven to create significant value for consumers and companies, but they pose a threat to consumers' privacy. A scenario-based experiment was conducted to uncover some of the underlying factors influencing the perceived quality of recommendations. The experiment finds no evidence to support a significant influence of transparency nor data control on the perceived recommendation quality. Further, this paper proposes that, in a non-privacy sensitive domain such as movie recommender, trust and privacy does not matter in the specific context.