What is Sentiment?
Sentiment checks whether a text is positive, negative, or neutral. It helps understand how people feel about a brand, product, or topic by analyzing their words and tone in social media posts, reviews, and comments.
Sentiment analysis in Brand24
You can quickly find mentions with a specific sentiment using our Sentiment filter. If you notice an error in sentiment identification, you can easily correct it manually by clicking on the sentiment within the mention.
Where to find Sentiment?
You can find Sentiment in almost all of our tabs!
How is Sentiment calculated?
Detecting sentiment in some posts can still be tricky, especially with sarcasm or complex context, but we're constantly improving!
Our sentiment analysis algorithm leverages the latest advantages of Machine Learning and Artificial Intelligence in general, namely Deep Learning and Pretrained Language Models (PLM), which are also used by companies like Google, Microsoft, Facebook, and Baidu. This results in the AI capabilities being as close as possible to the real understanding of mentions with the current state of scientific knowledge.
If you spot any errors, please let us know so we can keep improving and serve you better!
Additional features
For a deeper dive, explore our other complementary features, such as:
Emoji analysis
Intent filter
Emotion analysis
Context of a discussion
FAQ
Which languages does Sentiment Analysis support?
Which languages does Sentiment Analysis support?
Our sentiment analysis can process all languages. While the system is designed to handle a wide range of languages, the accuracy and performance may vary depending on the specific language and the complexity of the text.
Do you analyze Sentiment for all the mentions?
Do you analyze Sentiment for all the mentions?
Yes, though some analyses may be more reliable than others, depending on the language's popularity and the complexity of the text.
What quality metrics do you use?
What quality metrics do you use?
We use the F1-score as our main quality metric. The F1-score is 1, and the better the model performs. You can trust that this is one of the best quality measures available.
Thanks to the macro-averaged F1 metric, evaluated on 50k mentions, our Brand24 Sentiment Identification Algorithm has improved its accuracy from 61% to an impressive 95%.
We hope you found this article helpful! If you have any more questions or need assistance, feel free to reach out.