-2

I need to build a classifier which will classify any tweet string into a specific category. I've around 15 or so categories. I've the training dataset with me. Any ideas how should I go ahead with this? Using Python or Java for this.

I've been looking into Textblob and Stanford NLP classifiers. I'm just not sure which one to pick to get better results.

90abyss
  • 7,037
  • 19
  • 63
  • 94

1 Answers1

1

I enjoy using scikit-learn for supervised machine learning in Python. http://scikit-learn.org/stable/.

For Java, you might look into DKPro https://dkpro.github.io/, as they've leveraged existing toolkits for various NLP tasks.

And finally, whatever toolkit you use, you might look into whether you can use TweetNLP http://www.ark.cs.cmu.edu/TweetNLP/ for the POS tagging component.

Good luck!

jgloves
  • 719
  • 4
  • 14