DCN^2: Interplay of Implicit Collision Weights and Explicit Cross Layers for Large-Scale Recommendation

TL;DR

This paper introduces DCN^2, an improved version of the Deep and Cross network architecture, with algorithmic enhancements that improve modeling capabilities and efficiency in large-scale recommender systems, outperforming previous models.

Contribution

The paper presents three novel algorithmic improvements to DCNv2, enhancing its ability to model interactions and manage collisions, leading to superior performance in real-world and benchmark scenarios.

Findings

DCN^2 outperforms DCNv2 in live recommender systems.

The improvements address information loss and collision management issues.

DCN^2 achieves better results on benchmark datasets.

Abstract

The Deep and Cross architecture (DCNv2) is a robust production baseline and is integral to numerous real-life recommender systems. Its inherent efficiency and ability to model interactions often result in models that are both simpler and highly competitive compared to more computationally demanding alternatives, such as Deep FFMs. In this work, we introduce three significant algorithmic improvements to the DCNv2 architecture, detailing their formulation and behavior at scale. The enhanced architecture we refer to as DCN^2 is actively used in a live recommender system, processing over 0.5 billion predictions per second across diverse use cases where it out-performed DCNv2, both offline and online (ab tests). These improvements effectively address key limitations observed in the DCNv2, including information loss in Cross layers, implicit management of collisions through learnable lookup-level weights, and explicit modeling of pairwise similarities with a custom layer that emulates FFMs' behavior. The superior performance of DCN^2 is also demonstrated on four publicly available benchmark data sets.