Details
| Original language | German |
|---|---|
| Title of host publication | EuroVis 2019 |
| Subtitle of host publication | 21st Eurographics Conference on Visualization |
| Editors | Jimmy Johansson, Filip Sadlo, Elisabeta G. Marai |
| Pages | 85-89 |
| Number of pages | 5 |
| ISBN (electronic) | 978-3-03868-090-1 |
| Publication status | Published - 2019 |
| Event | 21st EuroVis 2019 - Porto, Portugal Duration: 3 Jun 2019 → 7 Jun 2019 |
Abstract
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
EuroVis 2019: 21st Eurographics Conference on Visualization. ed. / Jimmy Johansson; Filip Sadlo; Elisabeta G. Marai. 2019. p. 85-89.
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Online Learning of Visualization Preferences through Dueling Bandits for Enhancing Visualization Recommendations
AU - Rohs, Michael
AU - Kassel, Jan Frederik
N1 - Funding Information: ny opinions, findings, and conclusions expressed in this paper do not necessarily reflect the views of the Volkswagen Group.
PY - 2019
Y1 - 2019
N2 - A visualization recommender supports the user through automatic visualization generation. While previous contributions primarily concentrated on integrating visualization design knowledge either explicitly or implicitly, they mostly do not consider the user’s individual preferences. In order to close this gap we explore online learning of visualization preferences through dueling bandits. Additionally, we consider this challenge from a usability perspective. Through a user study (N = 15), we empirically evaluate not only the bandit’s performance in terms of both effectively learning preferences and properly predicting visualizations (satisfaction regarding the last prediction: µ = 85%), but also the participants’ effort with respect to the learning procedure (e.g., NASA-TLX = 24.26). While our findings affirm the applicability of dueling bandits, they further provide insights on both the needed training time in order to achieve a usability-aligned procedure and the generalizability of the learned preferences. Finally, we point out a potential integration into a recommender system.
AB - A visualization recommender supports the user through automatic visualization generation. While previous contributions primarily concentrated on integrating visualization design knowledge either explicitly or implicitly, they mostly do not consider the user’s individual preferences. In order to close this gap we explore online learning of visualization preferences through dueling bandits. Additionally, we consider this challenge from a usability perspective. Through a user study (N = 15), we empirically evaluate not only the bandit’s performance in terms of both effectively learning preferences and properly predicting visualizations (satisfaction regarding the last prediction: µ = 85%), but also the participants’ effort with respect to the learning procedure (e.g., NASA-TLX = 24.26). While our findings affirm the applicability of dueling bandits, they further provide insights on both the needed training time in order to achieve a usability-aligned procedure and the generalizability of the learned preferences. Finally, we point out a potential integration into a recommender system.
U2 - 10.2312/evs.20191175
DO - 10.2312/evs.20191175
M3 - Aufsatz in Konferenzband
SP - 85
EP - 89
BT - EuroVis 2019
A2 - Johansson, Jimmy
A2 - Sadlo, Filip
A2 - Marai, Elisabeta G.
T2 - 21st EuroVis 2019
Y2 - 3 June 2019 through 7 June 2019
ER -