This is the code and data for Mining the Past with Dual Criteria: Integrating Three types of Historical Information for Context-aware Event Forecasting

The code is tested to be runnable under the environment with python=3.10.4; pytorch=2.1.0; cuda=12.2; dgl=2.3.0+cu121 ...

The Datasets files need to be unzipped first:

unzip data.zip
unzip data_disentangled.zip

The folder ./data contains the original event data and the structured data.

where d_name is EG or IR or IS.

```
cd get_history

python get_back_embed.py  --c d_name
python get_graph_current.py  --datapath  ../data/d_name
python get_history_dict.py  


```

python ITHI.py --dataset d_name

python test.py --dataset d_name