ClardRec uses counterfactual-style score comparisons to split item representations into query-related and general parts, and transfers the general parts to improve recommendation.
Tsang, Yuangang Pan, and Jingchao Su
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Counterfactual Learning-Driven Representation Disentanglement for Search-Enhanced Recommendation
ClardRec uses counterfactual-style score comparisons to split item representations into query-related and general parts, and transfers the general parts to improve recommendation.