Source code of KDD23 Paper "Modeling Dual Period-Varying Preferences for Takeaway Recommendation"
source: download MT-small data from https://drive.google.com/drive/folders/1k9q9X4lK-pnJq2PaK6YtqSOCPOSU-O4D?usp=drive_link
structure:
code
mygraph
conf.py
main.py
utils.py
MT-small
data format:
- Train_data
| user | res | itemlst | label | hour |
|---|---|---|---|---|
| 0 | 0 | [1] | 1 | 14 |
| 1 | 1 | [2, 249, 291, 143, 53, 317, 126] | 1 | 22 |
| 2 | 2 | [3, 195, 204, 468] | 1 | 12 |
- Valid data or test data. For each ground truth valid/test data, we randomly select 99 stores that the user has not interacted with, serving as negative samples. Records sharing the same 'sample num' indicate that they are either a single positive sample or the negative samples derived based on the positive sample.
| user | Res | itemlst | label | hour | sample_num |
|---|---|---|---|---|---|
| 7 | 7 | [8, 244] | 1 | 18 | 0 |
| 7 | 4054 | [2030, 1743, 385] | 0 | 18 | 0 |
| 7 | 2183 | [687, 381, 515, 669] | 0 | 18 | 0 |
cd code
python3 main.py
@inproceedings{zhang2023modeling,
author = {Zhang, Yuting and Wu, Yiqing and Le, Ran and Zhu, Yongchun and Zhuang, Fuzhen and Han, Ruidong and Li, Xiang and Lin, Wei and An, Zhulin and Xu, Yongjun},
title = {Modeling Dual Period-Varying Preferences for Takeaway Recommendation},
booktitle={the 29th SIGKDD conference on Knowledge Discovery and Data Mining (KDD 2023)},
pages = {5628–5638},
numpages = {11},
year={2023}
}