The Sentiment Score, or polarity, is a metric output by sentiment analysis models (e.g., Google Cloud NLP, IBM Watson NLU) that expresses the emotional charge of the text content. The score generally ranges from negative (e.g., -1 or below zero) to positive (e.g., 1 or above zero).

In dashboard visualizations, the sentiment score is often represented vertically against rank (horizontal axis). Sentiment scoring is applied to document content, titles, and meta descriptions, and the classification can be nuanced (e.g., slightly negative, very positive, neutral). This metric is crucial for understanding how publications are leaning in their coverage of a brand.