%0 Conference Paper %1 Guesmi_2021 %A Guesmi, Mouadh %A Chatti, Mohamed Amine %A Vorgerd, Laura %A Joarder, Shoeb %A Zumor, Shadi %A Sun, Yiqi %A Ji, Fangzheng %A Muslim, Arham %B Adjunct Proceedings of the 29th ACM Conference on User Modeling, Adaptation and Personalization %D 2021 %I ACM %K explanation exum2021 interactive-recommender open-user-model recommender transparency umap2021 %P 246-252 %R 10.1145/3450614.3464479 %T On-demand Personalized Explanation for Transparent Recommendation %U https://doi.org/10.1145%2F3450614.3464479 %X The literature on explainable recommendations is already rich. In this paper, we aim to shed light on an aspect that remains under-explored in this area of research, namely providing personalized explanations. To address this gap, we developed a transparent Recommendation and Interest Modeling Application (RIMA) that provides on-demand personalized explanations with varying levels of detail to meet the demands of different types of end-users. The results of a preliminary qualitative user study demonstrated potential benefits in terms of user satisfaction with the explainable recommender system. Our work would contribute to the literature on explainable recommendation by exploring the potential of on-demand personalized explanations, and contribute to the practice by offering suggestions for the design and appropriate use of personalized explanation interfaces in recommender systems.

@inproceedings{Guesmi_2021, abstract = {The literature on explainable recommendations is already rich. In this paper, we aim to shed light on an aspect that remains under-explored in this area of research, namely providing personalized explanations. To address this gap, we developed a transparent Recommendation and Interest Modeling Application (RIMA) that provides on-demand personalized explanations with varying levels of detail to meet the demands of different types of end-users. The results of a preliminary qualitative user study demonstrated potential benefits in terms of user satisfaction with the explainable recommender system. Our work would contribute to the literature on explainable recommendation by exploring the potential of on-demand personalized explanations, and contribute to the practice by offering suggestions for the design and appropriate use of personalized explanation interfaces in recommender systems. }, added-at = {2021-07-13T18:31:51.000+0200}, author = {Guesmi, Mouadh and Chatti, Mohamed Amine and Vorgerd, Laura and Joarder, Shoeb and Zumor, Shadi and Sun, Yiqi and Ji, Fangzheng and Muslim, Arham}, biburl = {https://www.bibsonomy.org/bibtex/21a5e97ce5d6c406a2865f533deb299d9/brusilovsky}, booktitle = {Adjunct Proceedings of the 29th {ACM} Conference on User Modeling, Adaptation and Personalization}, description = {On-demand Personalized Explanation for Transparent Recommendation | Adjunct Proceedings of the 29th ACM Conference on User Modeling, Adaptation and Personalization}, doi = {10.1145/3450614.3464479}, interhash = {1ae2aa8c6409b1826251e4b11ddac305}, intrahash = {1a5e97ce5d6c406a2865f533deb299d9}, keywords = {explanation exum2021 interactive-recommender open-user-model recommender transparency umap2021}, month = jun, pages = {246-252}, publisher = {{ACM}}, timestamp = {2022-01-16T21:57:01.000+0100}, title = {On-demand Personalized Explanation for Transparent Recommendation}, url = {https://doi.org/10.1145%2F3450614.3464479}, year = 2021 }