Sentiment Analysis using Naive Bayes

ikayunida123ikayunida123 Member Posts: 17 Contributor I
edited December 2018 in Help

Hello there!

I'm planning to do a project about sentiment analysis. In RapidMiner website, it stated that rapid miner using Gaussian probability densities to model the Attribute data ( https://docs.rapidminer.com/latest/studio/operators/modeling/predictive/bayesian/naive_bayes.html ). I'm actually wondering if I can apply Gaussian distribution in sentiment analysis because all the journals that I've read all use multinomial distribution. Is Gaussian distribution can be used in sentiment analysis or text mining?

Thank you!

Sorry for my poor English :smileyembarrassed:

Best Answer

  • MariusHelfMariusHelf Posts: 1,869   Unicorn
    Solution Accepted

    While Gaussian Naive Bayes may not always be the best choice for text classification, it should work well enough. Just give it a try and cross-validate your results. That's the best way to find out if an algorithms works for your use  case.

     

    For text classification/sentiment classification, I would also try a linear SVM if naive bayes does not deliver satisfying results.

Answers

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761   Unicorn

    Oooooo a @MariusHelf sighting...

    MariusHelfsgenzer
Sign In or Register to comment.