Graph Convolution for Multimodal Information Extraction from Visually Rich Documents

TL;DR

This paper presents a graph convolutional model that effectively combines visual and textual information for entity extraction in visually rich documents, outperforming traditional sequence-based models.

Contribution

Introduces a novel graph convolution approach that integrates visual and textual features for improved information extraction from VRDs.

Findings

Our method significantly outperforms BiLSTM-CRF baselines.

Ablation studies confirm the effectiveness of visual-textual integration.

The approach achieves superior results on two real-world datasets.

Abstract

Visually rich documents (VRDs) are ubiquitous in daily business and life. Examples are purchase receipts, insurance policy documents, custom declaration forms and so on. In VRDs, visual and layout information is critical for document understanding, and texts in such documents cannot be serialized into the one-dimensional sequence without losing information. Classic information extraction models such as BiLSTM-CRF typically operate on text sequences and do not incorporate visual features. In this paper, we introduce a graph convolution based model to combine textual and visual information presented in VRDs. Graph embeddings are trained to summarize the context of a text segment in the document, and further combined with text embeddings for entity extraction. Extensive experiments have been conducted to show that our method outperforms BiLSTM-CRF baselines by significant margins, on two real-world datasets. Additionally, ablation studies are also performed to evaluate the effectiveness of each component of our model.