Runtime grammar productions creation

I am interested in natural languge processing with python-django.

My project requires creation of grammar productions at runtime. That means, whenever i ask a question or write a sentence, basic nlp steps like pos-tagging should be done and get them added to grammar produc开发者_运维问答tions or any other structure ( even if the words are not present in grammar) , so that i can do the further chunking and extraction of different nouns, verbs ,etc separately from that grammar.

Please guide me the same.

NLTK is the most popular python framework for NLP processing as I know it can do POS tagging, deep parsing, and allow you to write extraction rules based on those outputs.