Trinity: A Scenario-Aware Recommendation Framework for Large-Scale Cold-Start Users
This addresses cold-start challenges for early-stage users in new scenarios, particularly in large-scale product transitions, though it appears incremental by combining existing components into a synergistic framework.
The paper tackles the problem of cold-start recommendations for large-scale users in new scenarios, proposing the Trinity framework that integrates feature engineering, model architecture, and stable updating, and demonstrates substantial improvements in offline and online experiments for a billion-user Microsoft product transition.
Early-stage users in a new scenario intensify cold-start challenges, yet prior works often address only parts of the problem through model architecture. Launching a new user experience to replace an established product involves sparse behavioral signals, low-engagement cohorts, and unstable model performance. We argue that effective recommendations require the synergistic integration of feature engineering, model architecture, and stable model updating. We propose Trinity, a framework embodying this principle. Trinity extracts valuable information from existing scenarios while ensuring predictive effectiveness and accuracy in the new scenario. In this paper, we showcase Trinity applied to a billion-user Microsoft product transition. Both offline and online experiments demonstrate that our framework achieves substantial improvements in addressing the combined challenge of new users in new scenarios.