Bot That Detects Insults

This page is being edited.

The bot is trying to discourage behaviour that violates the principles of constructive communication. Particularly it can detect simple insults in messages and ask to rephrase such messages. The attached function performs insult detection.

The function uses a wonderful framework for creating dialog systems of DeepPavlov

pip install deeppavlov

from deeppavlov import build_model, configs

CONFIG_PATH = '/home/pi/bot/deeppavlov/configs/classifiers/insults_kaggle_conv_bert.json'
model = build_model(CONFIG_PATH, download=True)


def insult_detect(text):
    
    if model([text]) == ['Insult']:
        return True
    return False
    
if __name__ == "__main__":
    while True:
        text = input('enter a question: ')
        print(insult_detect(text))

from deeppavlov import build_model, configs

CONFIG_PATH = '/home/pi/bot/deeppavlov/configs/classifiers/insults_kaggle_conv_bert.json'

model = build_model(CONFIG_PATH, download=True)

def insult_detect(text):

if model([text]) == ['Insult']:

return True

return False

if __name__ == "__main__":

while True:

text = input('enter a question: ')

print(insult_detect(text))

Leave a Reply Cancel reply