Experiments codes for the paper:
Dugang Liu, Mingkai He, Jinwei Luo, Jiangxu Lin, Meng Wang, Xiaolian Zhang, Weike Pan and Zhong Ming. User-Event Graph Embedding Learning for Context-Aware Recommendation. To appear in SIGKDD '22.
Please cite our SIGKDD '22 paper if you use our codes. Thanks!
python==3.6.9
tensorflow==1.15.3
faiss-gpu==1.7.2
Note that our implementation is based on GCM (Link), and you may need to follow its instructions to compile the evaluator first.
For UEG-EL, a Yelp-OH based example is,
python UEG_EL.py --dataset Yelp-OH --num_gcn_layers 2 --reg 1e-1 --decoder_type FM --adj_norm_type ls --num_negatives 4 --test_batch_size 50 --pretrain 0 --save_flag 1 --finetune 0 --grouping 0
For UEG-EL-V, a Yelp-OH based example is,
python UEG_EL.py --dataset Yelp-OH --num_gcn_layers 2 --reg 1e-1 --decoder_type FM --adj_norm_type ls --num_negatives 4 --test_batch_size 50 --pretrain 1 --pretrain_layer 2 --pretrain_reg 1e-1 --save_flag 0 --finetune 0 --grouping 1 --keep_rate 0.9
python UEG_EL.py --dataset Yelp-OH --num_gcn_layers 2 --reg 1e-1 --decoder_type FM --adj_norm_type ls --num_negatives 4 --test_batch_size 50 --pretrain 1 --pretrain_layer 2 --pretrain_reg 1e-1 --save_flag 0 --finetune 1 --grouping 0 --keep_rate 0.9 --lr 0.0001
If you have any issues or ideas, feel free to contact us (dugang.ldg@gmail.com).