Cross-domain recommender systems adopt different tech- niques to transfer learning from source domain to target domain in order to alleviate the sparsity problem and im- prove accuracy of recommendations. Traditional techniques require the two domains to be linked by shared character- istics associated to either users or items. In collaborative filtering (CF) this happens when the two domains have over- lapping users or item (at least partially). Recently, Li et al.  introduced codebook transfer (CBT), a cross-domain CF technique based on co-clustering, and presented experimen- tal results showing that CBT is able to transfer knowledge between non-overlapping domains. In this paper, we dis- prove these results and show that CBT does not transfer knowledge when source and target domains do not overlap.
|Titolo:||Cross-domain recommendations without overlapping data: Myth or reality?|
|Autori interni:||CREMONESI, PAOLO|
|Data di pubblicazione:||2014|
|Appare nelle tipologie:||04.1 Contributo in Atti di convegno|