Implementation details

The system is implemented using the following languages and technologies:

• Python
• XSLT
• CSS
• R (data visualisation)
• Bash

The main body of the system (the text segmentation module) is implemented as a collection of Python modules:

• cosinemeasure.py
• peakpick.py
• micaseparse.py
• segmentation.py

segmentation.py is a front-end to the text segmentation module itself, which allows convenient coordination and manipulation of the text segmentation system from the command line or from a batch script. It accepts command-line arguments which select the type of output information required and specify the location of the input file (or files, if a pre-processed topic-break location list is available) - see section 3.4 for a list of command line options accepted.