https://github.com/nusnlp/reacts

The official code of the "Just What You Desire: Constrained Timeline Summarization with Self-Reflection for Enhanced Relevance" paper
https://github.com/nusnlp/reacts

large-language-models natural-language-processing pytorch timeline-summarization

Last synced: 27 days ago
JSON representation

The official code of the "Just What You Desire: Constrained Timeline Summarization with Self-Reflection for Enhanced Relevance" paper

• Host: GitHub
• URL: https://github.com/nusnlp/reacts
• Owner: nusnlp
• License: gpl-3.0
• Created: 2024-12-20T07:10:52.000Z (about 1 month ago)
• Default Branch: main
• Last Pushed: 2024-12-24T05:48:48.000Z (about 1 month ago)
• Last Synced: 2024-12-24T06:33:16.579Z (about 1 month ago)
• Topics: large-language-models, natural-language-processing, pytorch, timeline-summarization
• Language: Python
• Homepage: https://arxiv.org/abs/2412.17408
• Size: 44.9 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.txt

Awesome Lists containing this project

README

        
## CREST and REACTS

This repository provides the data and code to perform constrained timeline summarization, as reported in this paper:

> Just What You Desire: Constrained Timeline Summarization with Self-Reflection for Enhanced Relevance 


> [Muhammad Reza Qorib](https://mrqorib.github.io/), [Qisheng Hu](https://openreview.net/profile?id=~Qisheng_Hu1) and [Hwee Tou Ng](https://www.comp.nus.edu.sg/~nght/). 


> 2025. The 39th Annual AAAI Conference on Artificial Intelligence (to appear) [[PDF](https://arxiv.org/abs/2412.17408)].

Please cite our paper if you use its source code or data.

### Dataset

We provide the proposed CREST dataset under the folder `data`. The data is derived from the [ENTITIES](https://github.com/complementizer/news-tls) dataset. We do not own the copyright of the articles and the timeline summaries. Please contact the respective data owners for usage other than evaluating models on the constrained timeline summarization task.

### Code

We provide the code for our method, REACTS, under the folder `code`.

### License

This repository is licensed under the GNU General Public License Version 3 (see [License](./LICENSE.txt)).