Explainable Integration of Social Media Background in a Dynamic Neural Recommender

Recommender systems nowadays are commonly deployed in e-commerce platforms to help customers making purchase decisions. Dynamic recommender considers not only static user-item interaction data, but the temporal information at time of recommendation. Previous researches have suggested incorporate social media as dynamic neural recommenders after transforming them into embeddings. While such an approach can potentially improve recommendation performance, effectiveness is difficult explain. In this article, we propose explainable method integrate a recommender. Our applies association rule mining, which generate human-understandable behavior patterns from and platforms. With real-world show that integration accuracy by up 14% while using same data. Moreover, explain positive cases examining relevant rules.