Code accompanying the paper "Opinion Market Model: Stemming Far-Right Opinion Spread using Positive Interventions" [Calderon, et al. '24].
This repo contains the implementation for the bushfire case study discussed in the paper.
The code is split into two version folders:
v1/: version with regularization on two levels
v2/: version with platform-specific structure regularization (see Appendix D)
Each folder contains the following:
opinion_resources_opt1.py : opinion volume (tier 1) model estimation
opinion_resources_opt2.py : opinion share (tier 2) model estimation
opinion_analysis_v2.py : generative model sampling and fit evaluation routines
opinion_resources_twolevel.py : two-level fitting and post-fitting calculations
opinion_main_varyT_2D.py: main routine to be invoked through CLI
Both dataset and code are distributed under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license. If you require a different license, please contact us at piogabrielle.b.calderon@student.uts.edu.au or Marian-Andrei@rizoiu.eu.