2. Install all the requirements by`pip install -r requirements.txt`;

3. Start preprocess by`bash run.sh en`,`en` can be replaced by`zh` or`ar`;

4. Enter`n` to get data divided by filelist, or enter`y` and`train/dev/test rate`(e.g.`0.8 0.1 0.1`) to get data divided by sentences;

5. Enter`y` to get transform the data into BIO-type format, the transformed data will be in`output/BIO/`

6. The final output will be in`output/`.

5. Enter`y` to get transform the data into BIO-type format, the transformed data will be in`output/BIO/`, each train (test or dev) data will transform into 4 BIO-style json files(`token`,`entity_BIO`,`event_trigger_BIO` and`event_argument_BIO`);

6. The final output will be indirectory`output/`.

The output will save separately in`output/`, each file can be loaded by`json.loads()`. After loading, the data will be in`python list` type, each line will be in`python dict` type: