On the Influence of Demographics on Linguistic Alignment

MBWhitestone, Csirika

This repository contains code created for our computational dialogue modelling project at the University of Amsterdam. It provides free software to inspect the British National Corpus 2014 with the ConversAtion level Syntax SImilarity Metric (CASSIM) and Linguistic Inquiry and Word Count (LIWC) conversational alignment metric.

• cassim_inspect.ipynb: notebook for inspecting the output of cassim_run.py.
• cassim_run.py: code to run cassim on Conversations.
• cassim.py: a modified version of the CASSIM metric.
• conversations.py: converts the BNC2014 to Conversation classes.
• LICENSE: all our software is released under MIT. Software in cassim.py is released under the GNU General Public License v2.0.
• LIWC.py: code to run the LIWC metric on BNC2014.
• nlp_server.py: run this before and after running cassim_run.py; it will either start or stop the CoreNLP server.
• readme.md: this file containing important information.
• corenlp: this folder should contain an unpacked version of CoreNLP.
• data: this folder should contain an unpacked version of the BNC2014.

• Python3
• Java