python: modify PerceptronTagger in nltk to recognize 'and/or'

Question

How can I modify PerceptronTagger in nltk module (or maybe add some temporary functionality to it) so that it recognizes 'and/or' as 'CC' tag?

Answer 1

If this is the only thing you want to change, the simplest solution is to just post-process the tagged text:

for sentence in tagged_sentences:
    for n, (word,tag) in enumerate(sentence):
        if word == 'and/or':
            sentence[n] = (word, "CC")

But if your question is the first step to "improving" the NLTK's tagger, you should take the long view and think about how you could build or install a better tagger. Take a look at the many links included in this answer.