Recommender systems benefit people’s daily lives at every moment. While considerable attentions have been drawn by performance in one-step recommendation and static user-item network, recommenders’ performance on temporally evolving networks remains unclear. To address this issue, this .paper firstl