sociotechnical-transparency-abm

This repository contains the core logic for the ABM used in Gausen, A., Guo, C. & Luk, W. An approach to sociotechnical transparency of social media algorithms using agent-based modelling. AI Ethics (2024). https://doi.org/10.1007/s43681-024-00527-1.

Note: the published results where run on a version of the code but the core logic should be the same.

Summary:

Main.cpp: contains the C++ code with the ABM

Run_abm.py: contains the python code with the parameter inputs, calibration, genetic algorithm and wrapper to run the C++ code.

Parameters:

population = 250
neighbours = 25
n_sims = 1
n_posts = 10

preject_list = [0.01, 0.05]
preshare_list = [0.01, 0.05]
ponline_list = [0.1, 0.1]
psigma_list = [0.001, 0.05]

weights_chrono = [0.1, 1.0]
weights_belief = [0.1, 1.0]
weights_pop = [0.1, 1.0]
weights_random = [0.1, 1.0]
n_threads = 8
init_infl = 0.1

n_sims = 1
n_pop = 10
n_bits = 4
n_iter = 5
r_cross = 0.9
r_mut = 0.1

Inputs and Outputs: Input: Need a CSV with real world or dummy data on multiple tweets.

Output: Runs the optimum probabilities for each case study, optimum weights for all case studies, distance between simulated and real data.

How to Run:

1. Clone repo

2. Navigate to cloned repository

3. Make sure you have dependencies (python, g++, etc)

4. In run_abm.py update [INPUT DATA] with the name of your csv.

5. Run command: python3 run_abm.py

Data:

Column headings must be in alphabetical order. Column Headings: [counts prop_tweets_inf_rw steps tweet_id]

Licensing and Attribution:

This repo is under a BSD 3-Clause license. For proper attribution when using this code in any publications or research outputs, please cite our paper:

@article{gausen2024approach,
  title={An approach to sociotechnical transparency of social media algorithms using agent-based modelling},
  author={Gausen, Anna and Guo, Ce and Luk, Wayne},
  journal={AI and Ethics},
  pages={1--19},
  year={2024},
  publisher={Springer}
}

Suggested In-text Citation: Gausen, A., Guo, C. & Luk, W. An approach to sociotechnical transparency of social media algorithms using agent-based modelling. AI Ethics (2024). https://doi.org/10.1007/s43681-024-00527-1