Sentiment Analysis

Also known as: Sentiment Mining, Opinion Mining, Emotional Tone Analysis

NLP technique that classifies text (comments, reviews, verbatims) as positive, negative, or neutral, and detects specific emotions.

Sentiment Analysis is a Natural Language Processing (NLP) technique that automatically classifies text—reviews, social media comments, survey responses, brand mentions—according to its emotional polarity: positive, negative, or neutral. The most advanced models also identify specific emotions (joy, anger, fear, surprise) and sentiment intensity.

In market research, sentiment analysis has applications in: brand reputation monitoring on social media, verbatim analysis in satisfaction surveys, tracking conversations about product categories, and measuring the impact of communication campaigns.

Key challenges include irony and sarcasm (difficult to detect), slang and regionalisms (especially important in LATAM where Spanish varies significantly between countries), and mixed texts with multiple sentiments.

Atlantia applies sentiment analysis as an additional analytical layer in its qualitative studies and brand tracking.