Hao Liu, Xiangxian Li, Wei Gai1(B), Yu Huang, Jingbo Zhou
This paper introduces a personalized user interface element recom- mendation system, in which the model can recommend personalized user inter- face elements by introducing user features and user evaluations in the offline training. Through experiments, we found that compared with common machine learning algorithms, the Field-aware Factorization Machine that introduced user feature intersections has achieved a better accuracy in the recommendation, which shows the advantages of introducing user features and feature intersections in the recommendation of interface elements.