Frameworks, Modeling and Simulations of Misinformation and Disinformation: A Systematic Literature Review

TL;DR

This systematic review maps the landscape of frameworks, models, and simulations for mis/disinformation up to 2023, using PRISMA to analyze 57 studies. It finds that intentionality is central to differentiating misinformation from disinformation, with epidemiology- and opinion-dynamics-inspired models dominating representations and belief-updating simulations being especially prevalent. Validation is uneven, with many studies lacking empirical corroboration, while real-world data and simulations are the primary validation modes where used; health and political domains are most studied. The study highlights open challenges in standardization, contextual variability, and education as countermeasures, and suggests future work incorporating GenAI/LLMs and serious games to strengthen understanding and mitigation of mis/disinformation.

Abstract

The prevalence of misinformation and disinformation poses a significant challenge in today's digital landscape. That is why several methods and tools are proposed to analyze and understand these phenomena from a scientific perspective. To assess how the mis/disinformation is being conceptualized and evaluated in the literature, this paper surveys the existing frameworks, models and simulations of mis/disinformation dynamics by performing a systematic literature review up to 2023. After applying the PRISMA methodology, 57 research papers are inspected to determine (1) the terminology and definitions of mis/disinformation, (2) the methods used to represent mis/disinformation, (3) the primary purpose beyond modeling and simulating mis/disinformation, (4) the context where the mis/disinformation is studied, and (5) the validation of the proposed methods for understanding mis/disinformation. The main findings reveal a consistent essence definition of misinformation and disinformation across studies, with intent as the key distinguishing factor. Research predominantly uses social frameworks, epidemiological models, and belief updating simulations. These studies aim to estimate the effectiveness of mis/disinformation, primarily in health and politics. The preferred validation strategy is to compare methods with real-world data and statistics. Finally, this paper identifies current trends and open challenges in the mis/disinformation research field, providing recommendations for future work agenda.

Figures (9)

• Figure 1: PRISMA flow diagram with the following methodology.
• Figure 2: Research questions related to frameworks, modeling and simulation of mis/disinformation.
• Figure 3: Distribution of keywords and rate across the studies included in the systematic review.
• Figure 4: Number of selected studies and rate in the collection per year of publication.
• Figure 5: Paper distribution based on the definition provided.