Hateful, abusive, and profane content detection

The Watson Natural Language Processing HAP detection identifies hateful, abusive, and profane content (HAP content) in texts.

Block name

classification_transformer_en_slate.38m.hap

Supported languages

The HAP Detection block is able to detect HAP in English texts only.

Capabilities

Use this block to detect hateful, abusive or profane content in texts.

Dependencies on other blocks

None

Code sample

import watson_nlp

# Load the HAP model
hap_model = watson_nlp.load('classification_transformer_en_slate.38m.hap')

# Run the HAP model and print the result
hap_prediction = hap_model.run('Spiders are disgusting.')
print(hap_prediction)

Output of the code sample

{
  "classes": [
    {
      "class_name": "has_HAP",
      "confidence": 0.9629989862442017
    },
    {
      "class_name": "no_HAP",
      "confidence": 0.03700101003050804
    }
  ],
  "producer_id": {
    "name": "Transformer-based Text Classifier",
    "version": "0.0.1"
  }
}

Parent topic: Watson Natural Language Processing task catalog