Modelling the influence of individuals' and network characteristics on dissemination of fake news in a social network
The viral spread of digital misinformation has become so severe that the World Economic Forum (2018) considers it among the main threats to human society. There is worldwide concern over false news and the possibility that it can influence political, economic, and social well-being (Törnberg, 2018). The scale and rapidity of sharing fake news and misinformation is having an impact on democratic processes. False news can drive the misallocation of resources during terror attacks and natural disasters, the misalignment of business investments, and can misinform elections (Vosoughi et al., 2018). In order to curtail the negative influence of the fake news as an evolving phenomenon, we should continuously strive to understand it better, and study the mechanisms underlying the rapid diffusion of fake news in social networks. Most of existing research on fake news and related phenomena focuses on the analysis of past events by examining the spread of topics in social networks. While the analysis of large datasets (e.g. 500 milion tweets in Yang and Leskovec, 2010) is able to provide significant insight and enable the development of statistical models, and we have an understanding of the cognitive biases influencing individuals spreading fake news, we lack models that would allow us develop and test new theories to explain and predict this complex social phenomenon using rules that are at work at the level of individuals.
We thus intend to fill the gap by developing a new, original ABM model to develop and test theories on robust rules that influence the dissemination of fake news in a social network at the level of individuals. Objective of the proposed research is to develop and test new theories on rules that influence the dissemination of fake news in a social network at the level of individuals, using a new, original ABM model. New theories will provide a better understanding of the fake news phenomenon, while the novel ABM model will facilitate understanding of the individual and social dynamics present in the social networks where fake news proliferate and allow us and other researchers develop and test new theories. We intend to develop a set of experiments to research the relationship between relative success (domination in news cycle) of fake news and a set of factors, i.e. individual and network characteristics, e.g. cognitive biases, political bias, connectedness, fact-checking time, presence of hubs or ‘influencer’ nodes and echo chambers. We intend to use existing research on the fake news phenomenon and agent based modelling to develop and test the theories, and validate them by comparing model results and large datasets from main social network and news websites.
https://cris.cobiss.net/ecris/si/en/project/18758
The purpose of the simulation model is to model the dissemination of messages in a model of a social network. The model allows the selection of several synthetic social network models and a model based on the graph of a large component isolated from the Twitter data. The model allows the simulation of cascades occuring in social networks, and comparison of cascade properties depending on the network type, degree of the initial node in the cascade, and rules affecting the probability of message transmission on the level of individual nodes. Authors are available for consultation and help regarding the use of the model.
The model is built using Agent Based Modelling methodology in the Anylogic simulation modelling environment. Free version of the Anylogic environment is available from the publisher.
The simulation model is shared through the Creative Commons open-source license.
In the course of this research project we have collected approx. 100M Tweets posted between 2010 and 2022, with language code:"en", and containing hashtags or text with one of the search terms "Ukraine", "Ukraina", "UA". Due to changes in Twitter/X TOS in 2023 we are regrettably not allowed to share full data. List of Tweet IDs is available on request.
-
Preprint of upcoming paper on "Social Media Bot Detection Research": Review of Literature: https://arxiv.org/abs/2503.22838
-
»The Role of social networks in the spread of fake news«, scientific conference paper
- paper available at: https://press.um.si/index.php/ump/catalog/view/860/1251/3946, available in proceedings: https://press.um.si/index.php/ump/catalog/book/860
-
»The false information epidemic in social media : overview of recent developments and research«, scientific conference paper
- available in proceedings: https://www.fuds.si/wp-content/uploads/2025/02/SSSC-2025-book-of-abstract_final.pdf
-
»Modelling of Fake News Dissemination Using Agent Based Modelling«, invited lecture at VESS: Virtual EUROSIM Simulation Seminar, information available at: https://www.eurosim2022.eu/vess/#pastVESS
-
»Social networks, cognitive biases and fake news : can a simple computer model explain a complex phenomenon?«, scientific conference paper
- available in conference proceedings: https://www.itema-conference.com/wp-content/uploads/2021/04/ITEMA-2020_Conference-Proceedings_FINAL.pdf
-
Complete bibliography of research papers reviewed in the research project, APA 7 format.