Abstract | ||
---|---|---|
•TagMF, a recommendation method that enhances a latent factor model with additional information is presented.•The integrated model allows to provide richer interaction possibilities in Recommender Systems and to improve their transparency.•Users can express preferences through easily comprehensible tags instead of rating items, even in cold-start situations.•The method elucidates the hidden semantics of the factors and contributes to explaining a user’s preference profile.•Two user studies confirm that additional information improves, among others, perceived recommendation quality, as well as user experience in general. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1016/j.ijhcs.2018.05.002 | International Journal of Human-Computer Studies |
Keywords | DocType | Volume |
Recommender Systems,Collaborative Filtering,Interactive recommending,Matrix Factorization,Tags,Empirical studies,Human factors,User experience,User interfaces,User profiles | Journal | 121 |
ISSN | Citations | PageRank |
1071-5819 | 0 | 0.34 |
References | Authors | |
0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Benedikt Loepp | 1 | 88 | 10.71 |
Tim Donkers | 2 | 41 | 6.15 |
Timm Kleemann | 3 | 1 | 3.05 |
Jürgen Ziegler | 4 | 1028 | 300.31 |