product_train.csv

session_train.csv

1、只基于ID序列的下一个物品推荐：难度系数0.8

==2、基于ID序列和单语言的下一个物品推荐：难度系数1.0==

3、基于ID序列和多语言的下一个物品推荐：难度系数1.2

4、基于ID序列和单语言的下一个标题预测：难度系数1.2

5、基于ID序列和多语言的下一个标题预测：难度系数1.4

1. Make directory data under the project root. For instance,

   

2. run command pip install -r requirements.txt Tested under Python3.9/3.10, CUDA 11.7 CUDNN8 are required for Tensorflow.

cd preprocessing/

1. run preprocessing.py may include warnings but will not impact the overall performance.
2. run processing_nvt.py

cd model/

• For Tensorflow, run train.py
• For Pytorch, run train_torch.py