# An Empirical Survey on Long Document Summarization: Datasets, Models and Metrics

**Authors:** Huan Yee Koh, Jiaxin Ju, Ming Liu, Shirui Pan  
**arXiv:** [2207.00939](https://arxiv.org/abs/2207.00939)  
**Date:** 2022-07-03  
**Summary page:** https://www.summarizepaper.com/en/paper/2207.00939v1/

## Key points

- Automatic text summarization systems are important for condensing lengthy documents like academic articles and business reports.
- Recent advancements in the field include the integration of neural architectures.
- Challenges exist in extending these systems to long documents, highlighting the need for further research.
- The survey provides an overview of research on long document summarization, covering benchmark datasets, models, and evaluation metrics.
- The authors conduct an empirical analysis to assess current progress and suggest potential directions for future exploration.
- This survey is a valuable resource for researchers and practitioners interested in automatic text summarization for long documents.

## Summary

In their paper titled "An Empirical Survey on Long Document Summarization: Datasets, Models and Metrics," authors Huan Yee Koh, Jiaxin Ju, Ming Liu, and Shirui Pan explore the significance of automatic text summarization systems for condensing lengthy documents such as academic articles and business reports. These systems play a crucial role in aiding readers' comprehension by extracting key information. The authors highlight recent advancements in this field, particularly with the integration of neural architectures. They also discuss challenges associated with extending these systems to long documents and emphasize the need for further research. This survey provides a comprehensive overview of research on long document summarization, focusing on benchmark datasets, summarization models, and evaluation metrics. The authors organize existing literature within this context and conduct an empirical analysis to assess current progress. Their findings suggest potential directions for future exploration to advance this technology. This survey serves as a valuable resource for researchers and practitioners interested in automatic text summarization for long documents.