Sentiment analysis: Sentiment analysis is the use of AI to classify the emotional tone of text, typically as positive, negative, or neutral, and often in finer categories such as frustrated, satisfied, or urgent, across large volumes of reviews, tickets, or messages.
Older sentiment tools counted positive and negative words. Modern language models read context, so they can tell that "great, another outage" is not praise, that a polite complaint is still a complaint, and that a message about two topics can be positive about one and negative about the other. That last case, called aspect-based sentiment, is where most business value lives.
Typical uses: tagging support tickets by tone so angry customers get a faster response, tracking how reviews shift after a product change, reading open-ended survey answers at scale, and monitoring social mentions. You can run it in a spreadsheet by pasting text into a column and asking an assistant to classify each row, or automate it with a workflow that tags messages as they arrive.
Limits: sarcasm, industry jargon, and cultural differences still trip models up, and a model's "neutral" may hide a customer who is quietly leaving. Scores are only as consistent as your definitions, so give the model examples of each category from your own data. Always validate on a sample by hand before you trust a trend line, and never let sentiment scores alone trigger actions that affect a customer's account.
Example at work
A customer service manager exports 1,200 chat transcripts from the quarter and asks an assistant to tag each with sentiment and the main topic. The breakdown shows frustration clustering around a billing screen that engineering thought was fine. She reads thirty of the flagged transcripts to confirm before bringing it to the product team.
Why it matters
Sentiment analysis turns thousands of messages you could never read into patterns you can act on. Done carefully, with your own examples and a hand-checked sample, it is one of the quickest ways to find out what customers and employees actually think.
Related terms
- Natural language processing (NLP)Natural language processing (NLP) is the branch of AI focused on understanding and generating human language: reading, classifying, translating, summarizing, and answering questions in ordinary words. Large language models are its current dominant technology.
- SummarizationSummarization is the use of AI to condense a long document, transcript, thread, or dataset into a shorter version that keeps the key points. It is one of the most reliable everyday uses of language models, with known failure modes worth checking.
- Large language model (LLM)A large language model (LLM) is a neural network trained on enormous amounts of text to predict the next word, then tuned to follow instructions and hold a conversation. LLMs power ChatGPT, Claude, Gemini, and Microsoft Copilot.
- Workflow automationWorkflow automation connects apps and steps so that a trigger, such as a new form response or an incoming email, automatically runs a sequence of actions, increasingly with an AI step that reads, decides, or writes along the way.