---
license: cc-by-nc-4.0
language:
- en
tags:
- reddit
- cmv
size_categories:
- 100K<n<1M
---

# Basic Information
This dataset contains an enriched version of the [Winning Arguments Corpus](https://chenhaot.com/pages/changemyview.html).

The original corpus is enriched with
- Argument quality features obtained via [Adapter Models](https://github.com/Blubberli/ArgQualityAdapters)
- Linguistic features extracted with [elfen](https://elfen.readthedocs.io/en/latest/)

# Usage
To use this data, download the original corpus [here](https://chenhaot.com/pages/changemyview.html) and then merge on the `post_id`s and `comment_id`s.

Note that we only extract linguistic features for the *most upvoted*, *least upvoted*, and *delta* comment per post, and the post's text itself.
The argument quality features are available for all comments. If you need linguistic features for the rest of the comments, you can use [elfen](https://elfen.readthedocs.io/en/latest/) to extract them.

# Citation
If you use this enriched dataset in academic work, please cite

```bibtex
@inproceedings{doenmez-maurer-2025-ai,
title = "AI Argues Differently: Distinct Argumentative and Linguistic Patterns of LLMs in Persuasive Contexts",
author = "Dönmez, Esra and
Maurer, Maximilian and
Lapesa, Gabriella and
Falenska, Agnieszka",
year = {2025},
booktitle = "To appear: Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing",
}
```

and the original corpus
```
@inproceedings{tan+etal:16a,
author = {Chenhao Tan and Vlad Niculae and Cristian Danescu-Niculescu-Mizil and Lillian Lee},
title = {Winning Arguments: Interaction Dynamics and Persuasion Strategies in Good-faith Online Discussions},
year = {2016},
booktitle = {Proceedings of WWW}
}
```