Review-Based Cross-Domain Recommendation via Hyperbolic Embedding and Hierarchy-Aware Domain Disentanglement

arXiv:2403.20298v1 Announce Type: cross
Abstract: The issue of data sparsity poses a significant challenge to recommender systems. In response to this, algorithms that leverage side information such as review texts have been proposed. Furthermore, Cross-Domain Recommendation (CDR), which captures domain-shareable knowledge and transfers it from a richer domain (source) to a sparser one (target), has received notable attention. Nevertheless, the majority of existing methodologies assume a Euclidean embedding space, encountering difficulties in accurately representing richer text information and managing complex …

abstract algorithms arxiv challenge cs.ir cs.lg data domain embedding information issue knowledge recommendation recommender systems review sparsity systems type via