Find us on facebookFind us on facebook

Sentiment analyzer for English language

In the text boxes below you can write text on English language and press button "Analyze Sentiment". Sentiment of text will be shown under the button.


Underlaying technology

Sentiment analyzer for English language is made using machine learning algorithm (Naive Bayes), custom made tokenizer that was doing stemming (using Porter stemming algorithm), negation handling (S. Das and M. Chen negation handling algorithm), punction stripping etc.

Training set of learning algorithm is made out of 1000 positive and 1000 negative movie reviews from IMDB, containing about 700 000 words. Accuracy should be about 85%.

For more information about Sentiment analyzer or API behind it, please contact us.

LiBRE! časopis Donate bitcoins for