Short Text Pre-training with Extended Token Classification for E-commerce Query Understanding

TL;DR

This paper introduces Extended Token Classification (ETC), a novel pre-training task for short text query understanding in e-commerce, improving upon traditional masked language models by extending input tokens rather than masking, leading to better intent inference.

Contribution

The paper proposes ETC, a new pre-training method tailored for short e-commerce queries, addressing the limitations of masked language models in such contexts.

Findings

ETC outperforms traditional MLM in query understanding tasks.

The approach improves semantic understanding of short search queries.

Experiments demonstrate enhanced performance in e-commerce search scenarios.

Abstract

E-commerce query understanding is the process of inferring the shopping intent of customers by extracting semantic meaning from their search queries. The recent progress of pre-trained masked language models (MLM) in natural language processing is extremely attractive for developing effective query understanding models. Specifically, MLM learns contextual text embedding via recovering the masked tokens in the sentences. Such a pre-training process relies on the sufficient contextual information. It is, however, less effective for search queries, which are usually short text. When applying masking to short search queries, most contextual information is lost and the intent of the search queries may be changed. To mitigate the above issues for MLM pre-training on search queries, we propose a novel pre-training task specifically designed for short text, called Extended Token Classification (ETC). Instead of masking the input text, our approach extends the input by inserting tokens via a generator network, and trains a discriminator to identify which tokens are inserted in the extended input. We conduct experiments in an E-commerce store to demonstrate the effectiveness of ETC.