---
license: cc-by-nc-4.0
task_categories:
- text-generation
tags:
- recommendation-system
- preference-optimization
- dpo
---

# DynamicPO-Data

This repository contains the processed datasets for the paper [DynamicPO: Dynamic Preference Optimization for Recommendation](https://huggingface.co/papers/2605.00327).

**Official Code**: [xingyuHuxingyu/DynamicPO](https://github.com/xingyuHuxingyu/DynamicPO)

## Dataset Summary
DynamicPO is a plug-and-play dynamic preference optimization framework for LLM-based recommender systems. This repository provides the processed data used to evaluate the framework, following the construction pipeline of prior works like [LLaRA](https://arxiv.org/abs/2312.02445) and [S-DPO](https://arxiv.org/abs/2406.09215).

The collection includes processed splits for:
- **LastFM**: Music recommendation data.
- **Goodreads**: Book recommendation data.
- **Steam**: Game recommendation data.

The data consists of processed splits for supervised fine-tuning (SFT), preference optimization (PO), and evaluation.

## Data Preparation
As noted in the official GitHub repository, the LastFM data can be extracted using the following command:

```bash
cd ./data
unzip lastfm-sft-cans20.zip
```

The processed splits for Goodreads and Steam are also available within this dataset release.

## Citation
If you find this work useful, please consider citing the following paper:

```bibtex
@article{hu2026dynamicpo,
  title={DynamicPO: Dynamic Preference Optimization for Recommendation},
  author={Hu, Xingyu and Zhang, Kai and Wu, Jiancan and Wang, Shuli and Wang, Chi and Chen, Wenshuai and Zhu, Yinhua and Wang, Haitao and Wang, Xingxing and Wang, Xiang},
  journal={arXiv preprint arXiv:2605.00327},
  year={2026}
}
```