Unleash the Power of Context: Enhancing Large-Scale Recommender Systems with Context-Based Prediction Models
This work addresses performance enhancement for commercial recommender systems, offering a scalable solution with broad implications for personalized recommendations, though it appears incremental as it builds on existing CTR prediction models.
The paper tackles the problem of improving large-scale recommender systems by introducing Context-Based Prediction Models, which predict user actions using only user and contextual features, and reports significant improvements in offline and online business metrics with minimal serving cost impact.
In this work, we introduce the notion of Context-Based Prediction Models. A Context-Based Prediction Model determines the probability of a user's action (such as a click or a conversion) solely by relying on user and contextual features, without considering any specific features of the item itself. We have identified numerous valuable applications for this modeling approach, including training an auxiliary context-based model to estimate click probability and incorporating its prediction as a feature in CTR prediction models. Our experiments indicate that this enhancement brings significant improvements in offline and online business metrics while having minimal impact on the cost of serving. Overall, our work offers a simple and scalable, yet powerful approach for enhancing the performance of large-scale commercial recommender systems, with broad implications for the field of personalized recommendations.