Emotion Detection is a specialized module offered by the IBM Watson NLU API, going beyond standard sentiment analysis to detect specific emotional states. This module identifies specific emotions such as joy, fear, sadness, disgust, and anger expressed within the text.
Emotion analysis can be performed on customer review data and is typically executed programmatically (e.g., using a Google Colab notebook). The output includes specific numerical scores for each emotion (e.g., sadness score, joy score, fear score, disgust score, anger score). These scores are used to create custom fields and visualizations, such as a “Category Label (Emotion)” in Looker Studio.