PFCR

PyTorch implementation of the paper.

Lei Guo, Ziang Lu, Junliang Yu, Nguyen Quoc Viet Hung, Hongzhi Yin. Prompt-enhanced Federated Content Representation Learning for Cross-domain Recommendation. TheWebConf 2024.

Requirements

recbole==1.0.1
faiss-gpu==1.7.2
python==3.8.13
cudatoolkit==11.3.1
pytorch==1.11.0

Dataset

We use the processed datasets from UniSRec. You can download the processed dataset we used in the paper from 百度网盘.

Quick Start

Data Preparation

Preparing item codes:

python all_pq.py --gpu_id 0

Federated pre-train

python fed_pretrain.py

Before train, you need to modify the relevant configuration in the configuration files props/VQRec.yaml and props/pretrain.yaml.

Here are some important parameters in props/pretrain.yaml you may need to modify:

1.data_path: The path of the dataset you want to use for pre-training.

2.index_path: The path of the item codes you want to use for pre-training. It is best to maintain consistency with the data_path.

3.index_pretrain_dataset: Here, it needs to be set as the abbreviation of the first letter you use to pre train two datasets. For example, if you use "Office-Arts", please write it as OA here.

Fine-tune pre-trained recommender of "Office-Arts":

python single_train.py --d=OA --p=your_pretrained_model.pth

You also need to modify the index_pretrain_dataset in props/finetune.yaml to the abbreviation of the first letter of the current single dataset. The pq_data is consistent with the index_pretrain_dataset in props/pretrain.yaml.

Prompt-finetune

Prompt finetune pre-trained recommender of "Office-Arts":

python prompt_finetune.py --d=OA --p=your_pretrained_model.pth

You need to adjust the props/prompt.yaml in the same way as adjusting the props/finetune.yaml. In the model\, we also provide different prompt models for you to choose from.