Search papers, labs, and topics across Lattice.
This paper introduces the Orthogonal Disentanglement of Access habits (OrDA) framework, which addresses the issue of Pseudo-Positives in homepage marketing recommendations caused by habitual clicks masking content quality. By employing a dual-tower architecture with a gated allocation layer and orthogonal regularization, OrDA effectively separates interest signals from access habits, allowing for more accurate item ranking based solely on genuine user interest. Empirical evaluations reveal that OrDA significantly enhances click-through rates, achieving a 5.64% improvement on the Zhima homepage marketing block compared to existing methods.
OrDA's innovative approach eliminates access-habit bias, leading to a remarkable 5.64% increase in user click-through rates on marketing recommendations.
Clicks on homepage marketing blocks are driven by a dual-mechanism of content interest and access habits. However, habitual clicks often create Pseudo-Positives in marketing slots, where position advantage masks mediocre content quality, leading to biased recommendation ecosystems. We propose a framework called Orthogonal Disentanglement of Access habits (OrDA) to purify interest signals. OrDA utilizes a dual-tower structure with a gated allocation layer to adaptively route features and minimize interference. To ensure rigorous separation, we employ orthogonal regularization to constrain the latent interest and habit manifolds to be geometrically perpendicular. OrDA performs causal intervention (do-calculus) during inference to rank items solely by purified interest scores. Empirical online evaluations on large-scale datasets demonstrate that OrDA effectively eliminates access-habit bias, outperforming state-of-the-art methods in predictive accuracy. Online AB test 5.64% shows user click-through rates (UCTR) improvement on the Zhima homepage marketing block, Zhima rent-floor recommendation.
