Cooperative Retriever and Ranker in Deep Recommenders

TL;DR

This paper investigates the collaboration between retriever and ranker in deep recommender systems, addressing limitations of traditional training methods and proposing improved joint training strategies for better recommendation quality.

Contribution

It introduces novel methods for joint training of retriever and ranker, overcoming issues like distribution shift and false negatives in deep recommender systems.

Findings

Enhanced collaboration improves recommendation accuracy

Joint training reduces false negatives and distribution mismatch

Proposed methods outperform traditional pipelines

Abstract

Deep recommender systems (DRS) are intensively applied in modern web services. To deal with the massive web contents, DRS employs a two-stage workflow: retrieval and ranking, to generate its recommendation results. The retriever aims to select a small set of relevant candidates from the entire items with high efficiency; while the ranker, usually more precise but time-consuming, is supposed to further refine the best items from the retrieved candidates. Traditionally, the two components are trained either independently or within a simple cascading pipeline, which is prone to poor collaboration effect. Though some latest works suggested to train retriever and ranker jointly, there still exist many severe limitations: item distribution shift between training and inference, false negative, and misalignment of ranking order. As such, it remains to explore effective collaborations between retriever and ranker.