CrossDomain Recommendation Based on MetaData Using Graph Convolution Networks
Rabia Khan, Naima Iltaf, Rabia Latif, Nor Shahida Mohd Jamail · IEEE Access · 2023
Recent advancements in the domain of recommender systems have stemmed from the inspiration of representing the user-item interaction into graphs [11]. These heterogeneous graphs comprehensively capture the non-linear relationships between users and items alongwith features and emneddings. GCNs (Graph Convolution Networks) are state-of-the-art graph-based learning models that learn and represent the graph structures by recursively stacking layers of convolution and non-linear activation operations [3] [4]. GCNs are augmented with the strength of deep learning paradigms resulting in achieving better performance as compared to traditional CF (Collaborative Filtering) methods [7]. Despite modern improvements in the domain of recommender systems, cold-start users are a daunting challenge in the design of recommender systems since the conventional recommendation services are based on solely one data source [16]. During the recent years, cross-domain recommendation methods have gained popularity because of the availability of information in multiple domains for cold start users [34]. We supplement this information by utilizing the data contained in the metadata of users alongwith the strength of modelling graphs using GCNs [21]. Our proposed algorithm seams the strength of GCNs with cross domain paradigm utilizing the richness in metadata in user’s feedback to overcome the sparsity in user-item rating matrix. The combined advantages of GCNs and cross-domain approaches alleviated the issues of cold-start users by transferring user preferences from an source domain to a target domain [17].