Software

Here are a few basic scripts and tools for Natural Language Processing.

Toy Examples and Scripts

tf-idf: Example for converting documents into document vectors using tf-idf for weights
Document Similarity: Example for computing the pairwise document similarity between documents in a corpus

NL pipeline: The complete system that takes as input a text, parses it, converts it to logical form, and then derives inferences from it. The user needs to provide axioms in axioms.txt for deriving inferences from the system.
Lisp to XML: This software is a simple perl script that takes as input a lisp file and converts it to XML format. I have used this specifically to convert Charniak parse output into XML format. It requires the perl module called XML::TWIG to run.
Parse tree binarizer: This software is a perl script that takes as input an XML version of a parse tree, and converts it into a binary version. It requires the perl module XML::TWIG. This software is based on the Penn Treebank set of tags. For other tags, some rewrites need to be made to map the tags to the Penn Treebank form.

Mini-TACITUS: This software is an implementation of an abductive engine based on the paper by Jerry Hobbs. This software has been used successfully in previous projects (MOVER, LbR, Mobius-I, Mobius-II). I am currently working on extending this software to cover a large scale of axioms.

Logical Form Visualization Tool: This software is a visualization tool for LFToolkit which was written by Nishit Rathod. It requires Graphviz and perl to be pre-installed on your system.
Interpretation Visualization: This software helps in visualizing the best interpretation generated by TACITUS.

Charniak Parser for Windows: Several DARPA sponsored projects require executables to run on Windows machines. This is a standalone windows version of the Charniak Parser compiled using Ming32 libraries and Cygwin.