< Back to previous page

Project

Towards Tangible AI: An Alternative to Enhance the Experience with Personalized Recommendation Systems

Artificial Intelligence (AI) supports many of our everyday activities and decisions. However, personalized recommender systems (PRS) often produce adverse experiences due to a lack of awareness, control, or transparency. While research has directed solutions using graphical user interfaces (GUIs), there are no explorations of an alternative way to interact with these systems to improve the experience with such systems. Consequently, embedded in the Human-Computer Interaction (HCI) field, this thesis proposes a Design Science Research approach to explore the co-design of a Tangible User Interface (TUI) meant to improve the experience with PRS for videos/movies. In seven chapters including seven different studies, this thesis uncovers: (1) Methodological solutions to co-design the interaction with PRS considering the existent challenges on interfaces highly influenced by algorithmic systems; (2) A definition of the user experience with algorithmic systems such as PRS; (3) Common user beliefs of algorithmic systems such as PRS for videos/movies; (4) The characterization of the user experience with algorithmic systems such as PRS for videos/movies; (5) Existent tangible algorithmic imaginaries of algorithmic systems such as PRS for videos/movies; (6) The co-design of a TUI as an interaction alternative to improve the user experience of algorithmic systems such as PRS for videos/movies; and (7) Current research opportunities for TUIs meant to interact with algorithmic systems such as PRS in the area of group or collaborative recommendations. As a general conclusion, this thesis invites to reflect on the contributions and differences of concepts such as algorithmic beliefs and algorithmic imaginaries in this design context, some possible disadvantages of Tangible AI strategies, and the user perception towards Tangible AI. Additionally, it describes a set of limitations and considerations for this dissertation and presents future research opportunities to explore and impulse a more tangible way to interact with everyday algorithms.

Date:17 Jan 2018 →  26 Nov 2021
Keywords:Algorithmic experience, Algorithms, Digital Platforms, User experience
Disciplines:Human-computer interaction
Project type:PhD project