ESANS: Effective and Semantic-Aware Negative Sampling for Large-Scale Retrieval Systems

TL;DR

ESANS introduces a novel negative sampling method for large-scale retrieval systems, combining virtual sample generation and multimodal semantic clustering to improve diversity, reduce false negatives, and enhance retrieval performance.

Contribution

The paper presents ESANS, a new negative sampling approach that integrates EDIS and MSAC to address false negatives and semantic deficiencies in embedding-based retrieval.

Findings

ESANS improves retrieval accuracy in large-scale systems.

ESANS reduces false negatives through semantic-aware clustering.

Experiments show ESANS enhances efficiency and effectiveness.

Abstract

Industrial recommendation systems typically involve a two-stage process: retrieval and ranking, which aims to match users with millions of items. In the retrieval stage, classic embedding-based retrieval (EBR) methods depend on effective negative sampling techniques to enhance both performance and efficiency. However, existing techniques often suffer from false negatives, high cost for ensuring sampling quality and semantic information deficiency. To address these limitations, we propose Effective and Semantic-Aware Negative Sampling (ESANS), which integrates two key components: Effective Dense Interpolation Strategy (EDIS) and Multimodal Semantic-Aware Clustering (MSAC). EDIS generates virtual samples within the low-dimensional embedding space to improve the diversity and density of the sampling distribution while minimizing computational costs. MSAC refines the negative sampling distribution by hierarchically clustering item representations based on multimodal information (visual, textual, behavioral), ensuring semantic consistency and reducing false negatives. Extensive offline and online experiments demonstrate the superior efficiency and performance of ESANS.