sentiment analysis
In plain wordsUsing software to judge the feelings or opinions expressed in text — whether a review, post or comment is positive, negative or neutral, and about what.
The computational identification of opinions, evaluations and emotions expressed in text, most often as polarity classification (positive, negative, neutral) or a rating, at the level of documents, sentences or aspects — battery life negative, screen positive — in aspect-based analysis. Related tasks detect emotion categories, stance toward a target, subjectivity and sarcasm. Methods evolved from sentiment lexicons and hand-built rules through supervised classifiers trained on labeled reviews to fine-tuned transformers and prompted LLMs. It is used for brand and market monitoring, customer feedback, social-media analysis and finance. Difficulties include negation, irony, domain shift (a long battery life is good, a long wait is not) and variation across cultures and dialects, which can produce biased results.
Sentiment analysis flagged a spike in negative mentions after the update shipped.
Origin
Early work on classifying review polarity includes Peter Turney's unsupervised method and Bo Pang, Lillian Lee and Shivakumar Vaithyanathan's machine-learning study 'Thumbs up?' (both 2002). The name 'sentiment analysis' appears in Tetsuya Nasukawa and Jeonghee Yi's 2003 paper, and 'opinion mining' in Kushal Dave and colleagues' work of the same year.
Related
- text classificationtext classification, n.: Sorting pieces of text into categories automatically — what topic an article covers, whether a message is spam, which team should handle a customer's complaint.
- natural language processingnatural language processing, n.: The field that teaches computers to handle human language — to read, interpret, translate, summarize and produce text and speech. It is one of the oldest and largest areas of artificial intelligence.
- classificationclassification, n.: Sorting things into categories — spam or not spam, which animal is in a photo, which language a text is written in — by learning from examples that are already labeled.
- emotion recognition systememotion recognition system, n.: AI that tries to work out how people feel or what they intend from their faces, voices or bodies. The EU bans it at work and in schools and tightly regulates it elsewhere.
- algorithmic biasalgorithmic bias, n.: Unfairness in software that judges or serves people: it consistently produces worse results for some groups, such as women, older people or ethnic minorities, often because it learned from skewed or incomplete records.
Sources
- Pang, B., Lee, L. & Vaithyanathan, S. (2002). Thumbs up? Sentiment Classification using Machine Learning Techniques. EMNLP 2002. aclanthology.org/W02-1011
- Pang, B. & Lee, L. (2008). Opinion Mining and Sentiment Analysis. Foundations and Trends in Information Retrieval 2(1–2), 1–135.
Mentioned in
classificationclassification, n.: Sorting things into categories — spam or not spam, which animal is in a photo, which language a text is written in — by learning from examples that are already labeled., natural language processingnatural language processing, n.: The field that teaches computers to handle human language — to read, interpret, translate, summarize and produce text and speech. It is one of the oldest and largest areas of artificial intelligence., affective computingaffective computing, n.: Technology that tries to read people's emotions — from their faces, voices or words — or to display emotions of its own, like software that guesses your mood or a robot that seems to show feelings., text classificationtext classification, n.: Sorting pieces of text into categories automatically — what topic an article covers, whether a message is spam, which team should handle a customer's complaint.