At the moment, I am working on a project which requires me to build a naive Bayes classifier. Right now, I have a form online asking for people to submit a sentence and the subject of the sentence, in order to build a classifier to identify the subject of a sentence. But before I train the classifier, I intend on processing all entries for the parts-of-speech and the location of the subject. So my training set will be formatted as:
Sentence: Jake moved the chair. Subject: Jake
POS-Tagged: NNP VBD DD NN Location: 0
Would this be an effective way to build the classifier, or is there a better method?