NineRec Dataset Overview

Dataset Introduction

NineRec is a benchmark suite for evaluating transferable recommendation systems, published in TPAMI 2024. The suite supports multimodal, foundation‑model, transfer learning, and recommendation tasks.

Data Download

The complete dataset is publicly available via the following links:

• Google Drive: Source Dataset, Downstream Datasets

Data Format (Example: QB Dataset)

• QB_cover: Original images, filenames correspond to item IDs.
• QB_behaviour.tsv: User‑item interaction sequences; first column = user ID, second column = item ID sequence.
• QB_pair.csv: User‑item interaction pairs; columns: user ID, item ID, timestamp.
• QB_item.csv: Original text; columns: item ID, Chinese text, English text.
• QB_url.csv: Item URLs; columns: item ID, URL.

Citation

If you use this dataset, please cite:

@article{zhang2023ninerec,
      title={NineRec: A Benchmark Dataset Suite for Evaluating Transferable Recommendation},
      author={Jiaqi Zhang and Yu Cheng and Yongxin Ni and Yunzhu Pan and Zheng Yuan and Junchen Fu and Youhua Li and Jie Wang and Fajie Yuan},
      journal={arXiv preprint arXiv:2309.07705},
      year={2023}
}

Code Environment

• Pytorch==1.12.1
• cudatoolkit==11.2.1
• sklearn==1.2.0
• python==3.9.12

Data Preparation

Run get_lmdb.py to generate an LMDB database for image loading. Run get_behaviour.py to convert user‑item pairs into item‑sequence format.

Experiment Execution

Run train.py for pre‑training and transfer. Run test.py for evaluation.