Summary of Extract Information From Hybrid Long Documents Leveraging Llms: a Framework and Dataset, by Chongjian Yue et al.
Extract Information from Hybrid Long Documents Leveraging LLMs: A Framework and Dataset
by Chongjian Yue, Xinrun Xu, Xiaojun Ma, Lun Du, Zhiming Ding, Shi Han, Dongmei Zhang, Qi Zhang
First submitted to arxiv on: 28 Dec 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary Large Language Models (LLMs) excel in textual understanding and tabular reasoning tasks, but their performance on hybrid text containing both types of data remains unexplored. This paper addresses this gap by developing an Automated Information Extraction framework (AIE) to enable LLMs to process Hybrid Long Documents (HLDs). The AIE is applied to analyze four aspects of information extraction from HLDs: selecting and summarizing useful parts, serializing tables for understanding, adapting to complex scenarios with naive AIE, and enhancing LLMs through prompt engineering. Experimental findings highlight the importance of these approaches in processing HLDs. To support future research, a new dataset, Financial Reports Numerical Extraction (FINE), is proposed, along with publicly available code. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine trying to understand a long document that has both text and tables. This paper looks at how Large Language Models can handle this type of information. They developed a special tool called Automated Information Extraction to help these models process the documents. The researchers tested their approach in four different ways, finding out what works best for selecting important parts, converting tables into text, and more. They also created a new dataset with examples of financial reports that have both text and tables, so other researchers can use it too. |
Keywords
» Artificial intelligence » Prompt
