Note: this repository is depreciated, please use this repo

• Write new classifier that is scikit-learn compatible in src/classifier.
• Train the classifier with src/train_models.py and store it in model/

• For releasing, execute src/server.py
• For debugging, execute src/server.py -d

• For releasing, execute src/client_test.py
• For debugging, execute src/client_test.py -d

1. (GET) /litsmodel

Input:
None
Output:
{"model1_name":["emotion_A", "emotion_B", "emotion_C"], "model2_name":["emotion_D", "emotion_E"]}

2. (POST) /predict The output are the probability (or something similar) for each emotion belonging to the queried model name

Input:
{"model":"model1_name", "text": "whatever text here"}
Output:
{"res": [0.2, 0.3, 0.9]} 

3. (POST) /log

Input:
{"Whatever":"Whatever json object"}
Output:
None