QueryForm: A Simple Zero-shot Form Entity Query Framework

TL;DR

QueryForm is a novel zero-shot framework for extracting entities from form-like documents using a query-based approach with a dual prompting mechanism, achieving state-of-the-art results without target-specific training.

Contribution

It introduces a unified query-based framework with pre-training on large-scale weakly annotated data, enabling effective zero-shot document entity extraction across diverse layouts.

Findings

Achieves +4.6% to +10.1% F1 on XFUND benchmark

Achieves +3.2% to +9.5% F1 on Payment benchmark

Operates with smaller models and no image input

Abstract

Zero-shot transfer learning for document understanding is a crucial yet under-investigated scenario to help reduce the high cost involved in annotating document entities. We present a novel query-based framework, QueryForm, that extracts entity values from form-like documents in a zero-shot fashion. QueryForm contains a dual prompting mechanism that composes both the document schema and a specific entity type into a query, which is used to prompt a Transformer model to perform a single entity extraction task. Furthermore, we propose to leverage large-scale query-entity pairs generated from form-like webpages with weak HTML annotations to pre-train QueryForm. By unifying pre-training and fine-tuning into the same query-based framework, QueryForm enables models to learn from structured documents containing various entities and layouts, leading to better generalization to target document types without the need for target-specific training data. QueryForm sets new state-of-the-art average F1 score on both the XFUND (+4.6%~10.1%) and the Payment (+3.2%~9.5%) zero-shot benchmark, with a smaller model size and no additional image input.