Post-Click Conversion Rate (CVR) prediction aims to predict the probability of a conversion event occurring after a user clicks. Most CVR prediction methods use clicked events to train models and subsequently predict on both clicked and unclicked events, facing selection bias. To unbiasedly predict CVR, doubly robust (DR) learning incorporates propensity score reweighting and missing data error imputation, but with suboptimal performance under sparse click events. We theoretically demonstrate that existing DR methods face high or even unbounded bias, variance, and generalization error bound under small propensity scores from sparse click events. This motivates us to propose a new Batch-Adaptive DR (BADR) Learning method. In particular, we propose a BADR estimator, which adaptively adjusts the influence of each data batch during debiasing CVR prediction model training based on the propensity scores within that batch. We prove that the BADR estimator has bounded bias, variance, and generalization error bound, all of which are smaller than those of the DR estimator under small propensity scores, while maintaining asymptotic double robustness, i.e., achieving double robustness under a large sample size. Furthermore, we approximate the variance of the BADR estimator and derive a new batch-adaptive imputation model training loss compatible with the BADR estimator, which theoretically ensures further variance reduction during training. Our experiments on real-world datasets validate BADR’s effectiveness and rationality.

The post-click conversion rate (CVR) prediction task aims to predict the probability of a conversion after a click, which is essential in many fields. There are two widely-recognized challenges for CVR prediction: selection bias and data sparsity. Many previous methods focus on addressing selection bias by unbiasedly estimating the ideal loss based on the doubly robust estimator, which incorporates the error imputation model and propensity model to help CVR prediction model learning. However, they struggle with unreasonable knowledge transfer between the prediction model and imputation model and inflexible network structure design under sparse data. To this end, we introduce a novel principled adaptive structure learning approach, named Adap-SL, to adaptively learn the optimal network structure, adjust the number of activated (non-zero) parameters, and determine which knowledge needs to be transferred between the prediction model and the imputation model. Specifically, we start with an over-parameterized base network, where we adaptively extract partially overlapped subnetworks for the imputation model and the prediction model. Extensive experiments are conducted on three real-world recommendation datasets, demonstrating that our method consistently improves performance while requiring fewer parameters. The code is available at https://github.com/ChunyuanZheng/sigir25-sparse-sharing.

Existing Conversational Recommender Systems (CRS) predominantly utilize user simulators for training and evaluating recommendation policies. These simulators often oversimplify the complexity of user interactions by focusing solely on static item attributes, neglecting the rich, evolving preferences that characterize real-world user behavior. This limitation frequently leads to models that perform well in simulated environments but falter in actual deployment. Addressing these challenges, this paper introduces the Tri-Phase Offline Policy Learning-based Conversational Recommender System (TPCRS), which significantly reduces dependency on real-time interactions and mitigates overfitting issues prevalent in traditional approaches. TPCRS integrates a model-based offline learning strategy with a controllable user simulation that dynamically aligns with both personalized and evolving user preferences. Through comprehensive experiments, TPCRS demonstrates enhanced robustness, adaptability, and accuracy in recommendations, outperforming traditional CRS models in diverse user scenarios. This approach not only provides a more realistic evaluation environment but also facilitates a deeper understanding of user behavior dynamics, thereby refining the recommendation process.

Selection bias hinders recommendation models from learning unbiased user preference. Recent works empirically reveal that pursuing invariant user and item representation across biased and unbiased data is crucial for counteracting selection bias. However, our theoretical analysis reveals that simply optimizing representation invariance is insufficient for addressing the selection bias — recommendation performance is bounded by both representation invariance and discriminability. Worse still, current invariant representation learning methods in recommendation neglect even hurt the representation discriminability due to data sparsity and label shift. In this light, we propose a new Discriminative-Invariant Representation Learning framework for unbiased recommendation, which incorporates label-conditional clustering and prior-guided contrasting into conventional invariant representation learning to mitigate the impact of data sparsity and label shift, respectively. We conduct extensive experiments on three real-world datasets, validating the rationality and effectiveness of the proposed framework. Code and supplementary materials are available at: https://github.com/HungPaan/DIRL.