# Large Language Model for Table Processing: A Survey

**Authors:** Weizheng Lu, Jing Zhang, Ju Fan, Zihao Fu, Yueguo Chen, Xiaoyong Du  
**arXiv:** [2402.05121](https://arxiv.org/abs/2402.05121)  
**Date:** 2024-02-04  
**Summary page:** https://www.summarizepaper.com/en/paper/2402.05121v2/

## Key points

- Tables play a crucial role in organizing vast amounts of information for daily tasks
- Automation of tasks using Large Language Models (LLMs) and Visual Language Models (VLMs) has garnered significant interest
- Survey provides an overview of table-related tasks, including traditional tasks like table question answering and emerging fields like spreadsheet manipulation and table data analysis
- Training techniques tailored for LLMs and VLMs specifically designed for processing tables are covered
- Prompt engineering strategies are discussed to address challenges associated with processing implicit user intentions and extracting information from diverse table sources

## Summary

, , , , 


In the era of big data, tables play a crucial role in organizing vast amounts of information for various daily tasks. The automation of these tasks using Large Language Models (LLMs) or Visual Language Models (VLMs) has garnered significant interest from academia and industry. This survey provides an in-depth overview of table-related tasks, exploring both user scenarios and technical aspects. It delves into traditional tasks like table question answering as well as emerging fields like spreadsheet manipulation and table data analysis. The survey also covers training techniques tailored for LLMs and VLMs specifically designed for processing tables. Additionally, it discusses prompt engineering strategies, highlighting the challenges associated with processing implicit user intentions and extracting information from diverse table sources. : In the era of big data, tables play a crucial role in storing and organizing vast amounts of information. : Tables are essential for various daily tasks such as database queries, spreadsheet manipulations, web table question answering, and image table information extraction. : Automation of table-centric tasks using LLMs has garnered significant interest from academia and industry. : VLMs are also being utilized to automate table-related tasks. <kd> Automation of Tasks </ks>: The use of advanced language models is crucial in automating various table-related tasks for a wide range of applications.