Topic Identification with WEKA

Question

I am completely new to the field of Data mining and WEKA tool (just installed it today).

I need to do topic identification based on short text sentences.

Let say I have several categories: - politics - sports - other

I am thinking of doing the following: Have a list of terms that I compare the text to:

• Sports:
  • NFL
  • NBA
  • Touch down
  • etc
• Politics:
  • election
  • president
  • OBama
  • etc

Also, I would like to add more categories.

Then I would apply some algorithm SVM or Naive Bayes with the help of WEKA.

Any idea on how to start doing this with WEKA?

I have searched some tutorials on WEKA but I can't seem to get any examples similar to what I am trying to do.

Any help to start me up will be appreciated.