sentiment_analysis
Classify text sentiment as positive, negative, or neutral with a compound score and confidence. Handles negation, sarcasm, and mixed emotions in reviews, social posts, and support tickets.
Instructions
Analyse the emotional tone and sentiment of a piece of text.
Uses a fine-tuned language model to classify the overall sentiment and
return a continuous compound score. Handles negation, sarcasm signals,
and mixed-sentiment text. Works well on reviews, social posts, support
tickets, survey responses, and news excerpts.
Args:
text: Text to analyse. Optimal range: 10–2000 characters.
For longer documents, pass key paragraphs separately.
Returns:
label: Overall sentiment — 'positive', 'negative', or 'neutral'.
compound: Score from -1.0 (most negative) to +1.0 (most positive).
confidence: Model confidence in the classification, 0.0–1.0.
positive: Proportion of positive signal in the text.
negative: Proportion of negative signal in the text.
neutral: Proportion of neutral signal in the text.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| text | Yes |