DTDMN

Dynamic Dynamic Topic-Discourse Memory Networks (DTDMN) dynamically tracks the changes of latent topics and discourse in argumentative conversations, allowing the investigation of their roles in influencing the outcomes of persuasion for argumentation mining.

More details can be referred to:

What Changed Your Mind: The Roles of Dynamic Topics and Discourse in Argumentation Process. WebConf 2020.

Requirements

• Python >= 3.6
• Pytorch == 0.4.1

Data

CMV and Court dataset are in data/, with the preprocessing script process_json.py. Noted that you need to download the CMV dataset from here and extract to data/cmv/origin.

Take CMV for example, the preprocessing is as following:

$ cd data/cmv
$ python process_json.py

Usage

You can run the main code as:

$ python dtdmn_run.py