DisenCTR

This is the pytorch implementation for our SIGIR 2022 paper:

Yifan Wang, Yifang Qin, Fang Sun, Bo Zhang, Xuyang Hou, Ke Hu, Jia Cheng, Jun Lei, Ming Zhang (2022). DisenCTR: Dynamic Graph-based Disentangled Representation for Click-Through Rate Prediction

In this paper, we propose we construct a time-evolving user-item interaction graph induced by historical interactions. And based on the rich dynamics supplied by the graph, we propose a disentangled graph representation module to extract diverse user interests.

Please cite our paper if you use the code.

Environment Requirement

The code has been tested running under Python 3.8.13. The required packages are as follows:

• pytorch == 1.11.0
• torch_geometric == 2.0.4
• pandas == 1.4.1
• sklearn == 0.23.2

Running Example

For example, to generate ML-1M data for DisenCTR models, of whitch the raw data has been downloaded and unzipped at ~/raw_data/.

To process the data, run:

cd process_data
python process_ML.py

which will generate processed data files under the directory ~/Time_data/.

To conduct experiment on ML-1M, run:

cd ./code
python main.py --data ML --batch 1024 --patience 10 --nConvs 2 --K 4

For more execution arguments of DisenCTR, please refer to ~/code/main.py or run

python main.py -h