This is an implementation of Probability CKY parsing in python. The input grammar is as same as L1 grammar, which is shown in a sample grammar file pcfg.txt, and all the grammars will be binarized if it is needed.

The script takes two parameters:

• The grammar file
• The sentence to parse (which will be surrounded by double quotes)

For example: python prob-cky.py pcfg.txt "A test sentence ."

In the pre-processing part, I lemmatize some words that are not found in the grammar, so the nltk is included, which should be installed first.

• sudo pip install -U nltk