Nuvi uses a combination of human-annotated data and machine learning algorithms to put together a system that is more powerful than ever and can keep up with the ever-changing nuances of modern language.
Your monitors can now predict with more accuracy than ever whether an author of a mention was being positive, negative, or neutral with their message. Sentiment is one of the hardest facets of language analysis because it's subjective and can change based on context. We are confident that our Engine can keep up with the best of the best with its initial 80% accuracy because we have benchmarked it against the top solutions out there.
No sentiment engine can matter-of-factly tell whether a post is positive, neutral, or negative. Instead, the sentiment is analyzed on a scale. The Nuvi Language Engine assigns each word a specific value. Once a mention reaches a certain percentage threshold it is marked positive, neutral, or negative. The more positive a mention is, the brighter green the sentiment dot appears. The more negative, the redder. In the mention below, the sentiment score is 20. This means the Language Engine believes there is a 20% chance this mention is positive.
Check out these additional articles for more information:
Nuvi Language Engine: Everything You Need To Know
Nuvi Language Engine Data Breakdown
Comments
0 comments
Article is closed for comments.