Calibrating Video Watch-time Predictions with Credible Prototype Alignment

Accurately predicting user watch-time is crucial for enhancing user stickiness and retention in video recommendation systems. Existing watchtime prediction approaches typically involve transformations of watch-time labels for prediction and subsequent reversal, ignoring both the natural distribution properties of label and the instance representation confusion that results in inaccurate predictions. wo-stage method combining prototype learning and optimal transport for watch-time regression prediction, suitable for any deep recommendation model. Specifically, we observe that the watch-ratio (the ratio of watch-time to video duration) within same duration bucket exhibits a multimodal distribution. To facilitate incorporation into models, we use a hierarchical vector quantized variational autoencoder (HVQ-VAE) to convert the continuous label distribution into a high-dimensional discrete distribution, serving as credible prototypes for calibrations. Based on this, ProWTP views the alignment between prototypes and instance representations as a Semirelaxed Unbalanced Optimal Transport (SUOT) problem, where the marginal constraints of prototypes are relaxed. And the corresponding optimization problem is reformulated as a weighted Lasso problem for solution. Moreover, ProWTP introduces assignment and compactness losses to encourage instances to cluster closely around their respective prototypes, thereby enhancing the prototype-level distinguishability. Finally, we conducted extensive experiments, demonstrating our consistent superiority in real-world application.