0

I am learning the Semantic Role Labeling (SRL) task. I have read a lot, and now I come to a problem for how to represent the text features as vectors.

For example, for the sentence:

We like StackOverflow very much

given the predicate verb: like, a few features are:

the left 1st word: I
the right 1st word: StackOverflow
the POS tag of the left 1st word: Pronoun
The POS tag of the right 1st word: Adverbial

What are the right ways to represent these features as vectors?

If possible, can you also give me some guidances for how to normalize these features please?

I basically want to train the data with these type of features using SVM models.

Has QUIT--Anony-Mousse
  • 76,138
  • 12
  • 138
  • 194
leslie
  • 11,858
  • 7
  • 23
  • 22

1 Answers1

1

It doesn't matter what classifier you use (SVM or not) the feature generation for text is the same. I suggest you to take a look at this: Binary Feature Extraction

Also this library would make your life much easier: http://cogcomp.cs.illinois.edu/page/software_view/LBJ A tutorial is here: http://cogcomp.cs.illinois.edu/page/tutorial.201310

Community
  • 1
  • 1
Daniel
  • 5,839
  • 9
  • 46
  • 85