Towards Automated Neural Interaction Discovery for Click-Through Rate Prediction
This addresses the problem of improving personalized recommendations for billions of users in recommender systems, but it is incremental as it applies existing neural architecture search techniques to a specific domain.
The paper tackled the challenge of automating neural interaction discovery for click-through rate prediction by proposing AutoCTR, a framework that uses evolutionary architecture search with learning-to-rank guidance, achieving effectiveness and generalizability across datasets compared to human-crafted architectures.
Click-Through Rate (CTR) prediction is one of the most important machine learning tasks in recommender systems, driving personalized experience for billions of consumers. Neural architecture search (NAS), as an emerging field, has demonstrated its capabilities in discovering powerful neural network architectures, which motivates us to explore its potential for CTR predictions. Due to 1) diverse unstructured feature interactions, 2) heterogeneous feature space, and 3) high data volume and intrinsic data randomness, it is challenging to construct, search, and compare different architectures effectively for recommendation models. To address these challenges, we propose an automated interaction architecture discovering framework for CTR prediction named AutoCTR. Via modularizing simple yet representative interactions as virtual building blocks and wiring them into a space of direct acyclic graphs, AutoCTR performs evolutionary architecture exploration with learning-to-rank guidance at the architecture level and achieves acceleration using low-fidelity model. Empirical analysis demonstrates the effectiveness of AutoCTR on different datasets comparing to human-crafted architectures. The discovered architecture also enjoys generalizability and transferability among different datasets.