Automated Information Flow Selection for Multi-scenario Multi-task Recommendation

TL;DR

The paper tackles the challenge of multi-scenario multi-task recommendation by introducing AutoIFS, a lightweight framework that decouples four scenario-task information units using low-rank adaptation (LoRA) and learns to prune irrelevant information flows with a task-aware selection network. This adaptive flow pruning reduces noise and improves alignment of flows with specific scenarios and tasks, while maintaining a compact parameter footprint. Extensive experiments on two public datasets and online A/B testing demonstrate superior accuracy (AUC and log loss) and favorable efficiency compared with strong MSMTR baselines. The results underscore AutoIFS's practical potential for scalable, real-world MSMTR deployments with interpretable flow importance.

Abstract

Multi-scenario multi-task recommendation (MSMTR) systems must address recommendation demands across diverse scenarios while simultaneously optimizing multiple objectives, such as click-through rate and conversion rate. Existing MSMTR models typically consist of four information units: scenario-shared, scenario-specific, task-shared, and task-specific networks. These units interact to generate four types of relationship information flows, directed from scenario-shared or scenario-specific networks to task-shared or task-specific networks. However, these models face two main limitations: 1) They often rely on complex architectures, such as mixture-of-experts (MoE) networks, which increase the complexity of information fusion, model size, and training cost. 2) They extract all available information flows without filtering out irrelevant or even harmful content, introducing potential noise. Regarding these challenges, we propose a lightweight Automated Information Flow Selection (AutoIFS) framework for MSMTR. To tackle the first issue, AutoIFS incorporates low-rank adaptation (LoRA) to decouple the four information units, enabling more flexible and efficient information fusion with minimal parameter overhead. To address the second issue, AutoIFS introduces an information flow selection network that automatically filters out invalid scenario-task information flows based on model performance feedback. It employs a simple yet effective pruning function to eliminate useless information flows, thereby enhancing the impact of key relationships and improving model performance. Finally, we evaluate AutoIFS and confirm its effectiveness through extensive experiments on two public benchmark datasets and an online A/B test.

Figures (6)

• Figure 1: Overview of the multi-scenario multi-task recommendation framework, where arrows of different colors represent different relationship information flows.
• Figure 2: The architecture of the Automated Information Flow Selection (AutoIFS) framework, with Task 1 used as a specific task example, while the model actually outputs predictions for multiple tasks in parallel.
• Figure 3: Model size of our AutoIFS and baseline models on two datasets.
• Figure 4: Sensitivity analysis of the rank $r$, the final value $\gamma$ and the regularization penalty $\lambda$ on two datasets.
• Figure 5: Mask Visualization of our AutoIFS on two datasets.