Cross-domain Recommender Systems via Multimodal Domain Adaptation

Collaborative Filtering (CF) has emerged as one of the most prominent implementation strategies for building recommender systems. The key idea is to exploit the usage patterns of individuals to generate personalized recommendations. CF techniques, especially for newly launched platforms, often face a critical issue known as the data sparsity problem, which greatly limits their performance. Several approaches have been proposed in the literature to tackle the problem of data sparsity, among which cross-domain collaborative filtering (CDCF) has gained significant attention in the recent past. In order to compensate for the scarcity of available feedback in a target domain, the CDCF approach makes use of information available in other auxiliary domains. Most of the traditional CDCF approach aim is to find a common set of entities (users or items) across the domains and then use them as a bridge for knowledge transfer. However, most real-world datasets are collected from different domains, so they often lack information about anchor points or reference information for entity alignment. In this paper, we propose a domain adaptation technique to align the embeddings of users and items across the two domains. Our approach first exploits the available textual and visual information to independently learn a multi-view latent representation for each user and item in the auxiliary and target domains. The different representations of a user or item are then fused to generate the corresponding unified representation. A domain classifier is then trained to learn the embedding for the domain alignment by fixing the unified features as the anchor points. Experiments on two publicly benchmark datasets indicate the effectiveness of our proposed approach.