0

I am working an classifying tweets as positive or negative opinion. I heard Naive Bayes classifier is used commonly for classifying and I would like to write my own Naive Bayes Classifier for my project. Since, I am new to Data Mining and Naive Bayes classifier am not able to understand what and how to use Naive Bayes. Can anyone help me through??

merv
  • 67,214
  • 13
  • 180
  • 245
Madhan Gokul
  • 173
  • 1
  • 8

1 Answers1

1

First of all, a classifier needs to be trained with already classified data, the more the better; for example you need one hundred positive tweets and another one hundred negative.

If you're using python, you could start with NLTK and its huge collection of corpora

See also this question for a practical example.

Community
  • 1
  • 1
unixo
  • 548
  • 6
  • 16
  • 1
    it's merely a personal choice, also depending on your actual know-how. If you already use java, I'd suggest the great [Stanford University CoreNLP](http://nlp.stanford.edu/software/corenlp.shtml), entirely written in java; on the other hand, with python you could start with [NLTK](http://nltk.org/) or [pattern](http://www.clips.ua.ac.be/pattern). In my case, I preferred python over java (also because it's an interpreted language) – unixo Dec 30 '14 at 16:37