APRF-Net: Attentive Pseudo-Relevance Feedback Network for Query Categorization

TL;DR

This paper introduces APRF-Net, a neural network that improves categorization of rare e-commerce search queries by leveraging pseudo-relevance feedback from similar product documents, significantly outperforming existing models.

Contribution

The paper proposes a novel deep neural model, APRF-Net, that enhances rare query representation using pseudo-relevance feedback, addressing limitations of customer behavior data for infrequent queries.

Findings

APRF-Net improves query categorization F1@1 score by 5.9% over baselines.

The model achieves an 8.2% improvement for tail (rare) queries.

Experimental results demonstrate significant gains in search query understanding.

Abstract

Query categorization is an essential part of query intent understanding in e-commerce search. A common query categorization task is to select the relevant fine-grained product categories in a product taxonomy. For frequent queries, rich customer behavior (e.g., click-through data) can be used to infer the relevant product categories. However, for more rare queries, which cover a large volume of search traffic, relying solely on customer behavior may not suffice due to the lack of this signal. To improve categorization of rare queries, we adapt the Pseudo-Relevance Feedback (PRF) approach to utilize the latent knowledge embedded in semantically or lexically similar product documents to enrich the representation of the more rare queries. To this end, we propose a novel deep neural model named Attentive Pseudo Relevance Feedback Network (APRF-Net) to enhance the representation of rare queries for query categorization. To demonstrate the effectiveness of our approach, we collect search queries from a large commercial search engine, and compare APRF-Net to state-of-the-art deep learning models for text classification. Our results show that the APRF-Net significantly improves query categorization by 5.9% on F1@1 score over the baselines, which increases to 8.2% improvement for the rare (tail) queries. The findings of this paper can be leveraged for further improvements in search query representation and understanding.