Johannes Kunkel (former member)

Johannes Kunkel joined the Interactive Systems group in 2009. He started working as a student assistant and, after receiving his master’s degree in 2015, is now continuing the research around his master thesis as a doctoral researcher. His main research interest lies in Interactive Recommender Systems, Tangible User Interfaces and Gameful Design.

Interactive Recommender Systems: Typically Recommender Systems are not very transparent and lack efficient means to control them. In order to add interactivity to Recommender Systems, Johannes developed MovieLandscape, an application where users can interactively create a landscape to express their movie preferences and receive matching recommendations. MovieLandscape also constitutes the baseline for ongoing research. In late 2018 another application was developed following the same pattern: MusicPaths. While still under development, the application shows potential for letting users create personalized music playlists in an efficient and pleasurable way.

Tangible User Interfaces: Originating in a student project that was supervised by Johannes, a Tangible User Interface was created that combines a box filled with sand, a Microsoft Kinect depth sensor and a short throw projector. Later an application extended this AR-Sandbox, designed for helping visitors of historic museums to explore the exhibition. The system was published as a demo at MuC in 2017, where it won the Best Demo Award. Recently the projection of the AR-Sandbox was combined with the Microsoft HoloLens, creating 3D depictions of the historic city of Duisburg depending on the depth a user dug into the sand.

Gameful Design: One of the earliest developments, Johannes supervised, was Kinben – The Gamification Engine. Kinben is an open source project, implementing Gamification-as-a-Service and its source code is accessible on GitHub under the GNU Lesser General Public License. Later, Kinben was used for developing a playful Task Management System, which was tested live at offices of Centigrade in Mülheim an der Ruhr. Recently, Johannes’ research interest has shifted from Gamification to Games With a Purpose. MuchoMatcho is an online game for two players that helps tagging latent factors of Matrix Factorization. MuchoMatcho is currently subject of further development. In early 2019, LittleMissFits was developed. It is also a Game with a Purpose focusing latent factor semantics. But instead of tagging factors, players try to determine which movies belong to a factor and which do not. LittleMissFits is a single player game and can still be played online.

Thesis

Contact

Mail

Publications

NewsViz: Depicting and Controlling Preference Profiles Using Interactive Treemaps in News Recommender Systems

Exploring Mental Models for Transparent and Controllable Recommender Systems: A Qualitative Study

Let Me Explain: Impact of Personal and Impersonal Explanations on Trust in Recommender Systems

Ein kollaboratives Task-Management-System mit spielerischen Elementen

Understanding Latent Factors Using a GWAP

Trust-Related Effects of Expertise and Similarity Cues in Human-Generated Recommendations

Eine Sandbox zur physisch-virtuellen Exploration von Ausgrabungsstätten

A 3D Item Space Visualization for Presenting and Manipulating User Preferences in Collaborative Filtering

3D-Visualisierung zur Eingabe von Präferenzen in Empfehlungssystemen