Читать книгу Enterprise AI For Dummies - Zachary Jarvinen - Страница 43

Sentiment analysis uses text mining, NLP, and other AI techniques to detect the opinions and emotions of a person based on written or spoken content, such as social media posts, reviews, videos, and podcasts. It identifies the person expressing the opinion, what the person is talking about, and whether the opinion is positive or negative.

Also called opinion mining, sentiment analysis is often used to process reviews or survey results to discern the voice of the customer and adjust a policy, product, or response accordingly.

In the early days of Twitter, many corporate social media teams would auto-retweet any content that tagged the brand. The resulting retweets of complaints of bad service were a great source of amusement to the general populace but did little to build the value of the brand. Sentiment analysis not only avoids such embarrassing moments, but also creates an opportunity to be proactive in engaging customers with knowledge and empathy.