OptMSM: Optimizing Multi-Scenario Modeling for Click-Through Rate Prediction

Xing Tang,Yang Qiao,Yuwen Fu,Fuyuan Lyu,Dugang Liu,Xiuqiang He

2023-06-23

Abstract:A large-scale industrial recommendation platform typically consists of multiple associated scenarios, requiring a unified click-through rate (CTR) prediction model to serve them simultaneously. Existing approaches for multi-scenario CTR prediction generally consist of two main modules: i) a scenario-aware learning module that learns a set of multi-functional representations with scenario-shared and scenario-specific information from input features, and ii) a scenario-specific prediction module that serves each scenario based on these representations. However, most of these approaches primarily focus on improving the former module and neglect the latter module. This can result in challenges such as increased model parameter size, training difficulty, and performance bottlenecks for each scenario. To address these issues, we propose a novel framework called OptMSM (\textbf{Opt}imizing \textbf{M}ulti-\textbf{S}cenario \textbf{M}odeling). First, we introduce a simplified yet effective scenario-enhanced learning module to alleviate the aforementioned challenges. Specifically, we partition the input features into scenario-specific and scenario-shared features, which are mapped to specific information embedding encodings and a set of shared information embeddings, respectively. By imposing an orthogonality constraint on the shared information embeddings to facilitate the disentanglement of shared information corresponding to each scenario, we combine them with the specific information embeddings to obtain multi-functional representations. Second, we introduce a scenario-specific hypernetwork in the scenario-specific prediction module to capture interactions within each scenario more effectively, thereby alleviating the performance bottlenecks. Finally, we conduct extensive offline experiments and an online A/B test to demonstrate the effectiveness of OptMSM.

Information Retrieval

What problem does this paper attempt to address?

This paper attempts to address the issues present in multi-scenario Click-Through Rate (CTR) prediction on large-scale industrial recommendation platforms. Specifically, existing multi-scenario CTR prediction methods typically consist of two main modules: a scenario-aware learning module and a scenario-specific prediction module. Most existing methods primarily focus on improving the scenario-aware learning module to enhance the effectiveness of multi-functional representations, while neglecting improvements to the scenario-specific prediction module. This leads to problems such as increased model parameter size, higher training difficulty, and performance bottlenecks in each scenario.

To address these issues, the authors propose a new framework called OptMSM (Optimized Multi-Scenario Modeling). This framework mainly includes the following aspects:

1. **Simplified and Effective Scenario-Enhanced Learning Module**: The input features are divided into scenario-specific features and scenario-shared features, which are then mapped to specific information embeddings and a set of shared information embeddings, respectively. Orthogonal constraints are applied to promote the decoupling of shared information corresponding to each scenario, which is then combined with specific information embeddings to obtain multi-functional representations.

2. **Scenario-Specific Hypernetwork**: A scenario-specific hypernetwork is introduced in the scenario-specific prediction module to more effectively capture interactions within each scenario, thereby alleviating performance bottlenecks.

Through extensive offline experiments and online A/B testing, the authors demonstrate the effectiveness of OptMSM. Additionally, this framework can be effectively integrated with other existing multi-scenario models to further enhance performance.