Cold-start Sequential Recommendation via Meta Learner

This paper explores meta-learning in sequential recommendation to alleviate the item cold-start problem. Sequential aims capture user's dynamic preferences based on historical behavior sequences and acts as a key component of most online scenarios. However, previous methods have trouble recommending items, which are prevalent those As there is generally no side information task, could not be applied when only user-item interactions available. Thus, we propose Meta-learning-based Cold-Start Recommendation Framework, namely Mecos, mitigate problem recommendation. task non-trivial it targets at an important novel challenging context. Mecos effectively extracts user preference from limited learns match target with potential user. Besides, our framework can painlessly integrated neural network-based models. Extensive experiments conducted three real-world datasets verify superiority average improvement up 99%, 91%, 70% HR@10 over state-of-the-art baseline methods.