Deep Learning based Key Information Extraction from Business Documents: Systematic Literature Review

2024-07-23

arXiv.org

AI Summary

This study addresses the low efficiency and poor automation in key information extraction from commercial documents. We systematically review 96 deep learning methods published between 2017 and 2023, establishing the first comprehensive, rigorously categorized methodology framework for document understanding. A three-dimensional evaluation framework—assessing structural capability, generalization, and annotation efficiency—is proposed, revealing three core bottlenecks: weak cross-domain transferability, poor few-shot adaptability, and absence of end-to-end structured output. Integrating techniques from NLP, document image analysis, sequence modeling, and multimodality, we empirically evaluate mainstream models—including BERT, LayoutLM, Donut, and PPOCR—as well as the pretraining-fine-tuning paradigm. We open-source a reproducible classification matrix and benchmark suite. Our findings identify three priority research directions, providing both theoretical foundations and practical guidance for next-generation intelligent document processing systems.

Technology Category

• Natural Language Processing: Information Extraction
• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Large Vision Models

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Extracting key information from documents represents a large portion of business workloads and therefore offers a high potential for efficiency improvements and process automation. With recent advances in deep learning, a plethora of deep learning-based approaches for Key Information Extraction have been proposed under the umbrella term Document Understanding that enable the processing of complex business documents. The goal of this systematic literature review is an in-depth analysis of existing approaches in this domain and the identification of opportunities for further research. To this end, 96 approaches published between 2017 and 2023 are analyzed in this study.

Problem

• Extracting key information from business documents efficiently
• Analyzing Deep Learning approaches for Document Understanding
• Identifying research gaps in Key Information Extraction methods

Innovation

• Deep Learning for document information extraction
• Systematic review of 130 Document Understanding approaches
• Analysis of efficiency improvements in business processes