PythonLSA runs in Python, available for download here.
PythonLSA also requires the Numarray python package, available for download here.
Both are available as self-installing windows executables.
PythonLSA includes statistics code from Gary Strangman.
The main module is lsa_model.py. Sample transformation are in the file go.py.
size_splitter.py can be used to split large documents into equal-sized chunks.
splitter.py can be used to split large documents into chunks, delimeted by certain text. (For example, dividing a book into chapters based on the word "CHAPTER".)