Non-parametric Graph Convolution for Re-ranking in Recommendation Systems

TL;DR

This paper introduces a non-parametric graph convolution re-ranking method for recommendation systems that improves ranking quality during testing with minimal additional computational cost, addressing scalability issues of graph-based methods.

Contribution

It proposes a novel test-time only graph convolution re-ranking strategy that is plug-and-play and significantly reduces computational overhead in real-world RecSys.

Findings

Achieves an average of 8.1% improvement in ranking quality.

Maintains low additional computational cost of about 0.5.

Effective across multiple benchmark datasets with varying sparsity.

Abstract

Graph knowledge has been proven effective in enhancing item rankings in recommender systems (RecSys), particularly during the retrieval stage. However, its application in the ranking stage, especially when richer contextual information in user-item interactions is available, remains underexplored. A major challenge lies in the substantial computational cost associated with repeatedly retrieving neighborhood information from billions of items stored in distributed systems. This resource-intensive requirement makes it difficult to scale graph-based methods in practical RecSys. To bridge this gap, we first demonstrate that incorporating graphs in the ranking stage improves ranking qualities. Notably, while the improvement is evident, we show that the substantial computational overheads entailed by graphs are prohibitively expensive for real-world recommendations. In light of this, we propose a non-parametric strategy that utilizes graph convolution for re-ranking only during test time. Our strategy circumvents the notorious computational overheads from graph convolution during training, and utilizes structural knowledge hidden in graphs on-the-fly during testing. It can be used as a plug-and-play module and easily employed to enhance the ranking ability of various ranking layers of a real-world RecSys with significantly reduced computational overhead. Through comprehensive experiments across four benchmark datasets with varying levels of sparsity, we demonstrate that our strategy yields noticeable improvements (i.e., 8.1% on average) during testing time with little to no additional computational overheads (i.e., 0.5 on average). Code: https://github.com/zyouyang/RecSys2025_NonParamGC.git