Modelling the Closed Loop Dynamics Between a Social Media Recommender System and Users' Opinions

2025-07-26

AI Summary

This study investigates the bidirectional, closed-loop coupling between social media recommendation systems and user opinion dynamics: algorithms generate content recommendations based on historical user behavior, while user interactions reciprocally shape evolving opinions and subsequent recommendations. To formalize this feedback loop, we propose a unified mathematical model integrating opinion evolution with algorithmic recommendation feedback, analyzed quantitatively via Monte Carlo simulation. Key contributions include: (1) demonstrating that exposure to moderately positioned content induces only marginal opinion shifts; (2) showing that virality-aware recommendation strategies significantly mitigate opinion polarization; and (3) identifying critical opinion distribution intervals highly sensitive to polarization onset. This work is the first to quantify, within a single theoretical framework, how recommender systems actively drive both polarization and radicalization. Furthermore, it proposes actionable intervention strategies that jointly optimize platform engagement metrics and societal well-being.

Technology Category

• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Data Mining & Knowledge Management: Recommender Systems
• Machine Learning: Learning Preferences or Rankings

Application Category

• Social Networks and Social Media: Media and governance, opinion dynamics, filter bubbles, polarization
• User Modeling, Personalization and Recommendation: Social recommender systems and personalization for the social Web
• Web Mining and Content Analysis: Sentiment analysis and opinion mining

Abstract

This paper proposes a mathematical model to study the coupled dynamics of a Recommender System (RS) algorithm and content consumers (users). The model posits that a large population of users, each with an opinion, consumes personalised content recommended by the RS. The RS can select from a range of content to recommend, based on users' past engagement, while users can engage with the content (like, watch), and in doing so, users' opinions evolve. This occurs repeatedly to capture the endless content available for user consumption on social media. We employ a campaign of Monte Carlo simulations using this model to study how recommender systems influence users' opinions, and in turn how users' opinions shape the subsequent recommended content. We take an interest in both the performance of the RS (e.g., how users engage with the content) and the user's opinions, focusing on polarisation and radicalisation of opinions. We find that different opinion distributions are more susceptible to becoming polarised than others, many content stances are ineffective in changing user opinions, and creating viral content is an effective measure in combating polarisation of opinions.

Problem

• Modeling dynamic interaction between recommender systems and user opinions
• Studying polarization and radicalization effects from algorithmic recommendations
• Analyzing viral content impact on opinion polarization mitigation

Innovation

• Mathematical model for RS-user dynamics
• Monte Carlo simulations for opinion analysis
• Viral content reduces opinion polarisation