Abstract
Cross-domain recommendation is an effective technique to alleviate the data sparsity problem in recommender systems by utilizing the information from relevant domains. In this paper, we propose Cross-domain Deep Neural Network (CD-DNN) for the cross-domain recommendation. CD-DNN solves the rating prediction problem by modeling users and items using reviews and item metadata, which jointly learns features of users and items from not only the target domain but also other source domains. Latent factors for users and items are learned by several parallel neural networks, and the relevance of user features and item features is learned by maximizing prediction accuracy. CD-DNN builds a single mapping for user features in the latent space, so that the network for user is optimized together with item features from other domains. Experimental results indicate that the proposed CD-DNN significantly outperforms other state-of-the-art recommendation approaches on four public datasets of Amazon and it alleviates the data sparsity problem by leveraging more data across domains.
Highlights
IN the era of information explosion, information overload is one of the difficulties we face
We show in our experiments that Crossdomain Deep Neural Network (CD-DNN) significantly outperforms the current models and it alleviates the data sparsity problem by leveraging more data across domains
RELATED WORK we mainly summarize three categories of studies related to our work, including recommender systems using text data, deep learning based recommender systems, and cross-domain recommendation
Summary
IN the era of information explosion, information overload is one of the difficulties we face. The content-based recommendation method relies on extracting features from user preferences and items, and does not need too many rating records, which indicates that there is almost no problem of data sparsity. If apply it to recommendations for new users, the accuracy of recommendations may not be good enough, because in this case the user profiles are too difficult to obtain. A common solution to the data sparsity problem is to integrate collaborative filtering with content information to form a hybrid approach because users and items are generally associated with content information like reviews or item metadata. We propose a cross-domain deep neural network (CD-DNN) to jointly learn features of users and items from different domains with reviews and item metadata for rating prediction problems.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
