Hierarchical Multi-field Representations for Two-Stage E-commerce Retrieval

Niklas Freymuth, Dong Liu, Thomas Ricatte, Saab Mansour

Keywords: dense retrieval, document representation, re-ranking, contrastive learning

Abstract:

Dense retrieval methods typically target unstructured text data represented as flat strings. However, e-commerce
catalogs often include structured information across multiple fields, such as brand, title, and description, which
contain important information potential for retrieval systems. We present the Cascading Hierarchical Attention
Retrieval Model (CHARM), a novel framework designed to encode structured product data into hierarchical
field-level representations with progressively finer detail. Utilizing a novel block-triangular attention mechanism,
our method captures the inter-dependencies between product fields in a hierarchical manner, yielding field-level
representations and aggregated vectors suitable for fast and efficient retrieval. Combining both representations
enables a two-stage retrieval pipeline, in which the aggregated vectors support initial candidate selection, while
more expressive field-level representations facilitate precise fine-tuning for downstream ranking. Experimentally,
CHARM provides higher-quality retrieval compared to state-of-the-art dense retrieval methods on publicly available
large-scale e-commerce datasets. Further, our analysis highlights the framework’s ability to align different queries
with appropriate product fields, enhancing retrieval accuracy and explainability.

Submission Number: 2