Why am I seeing this? Towards recognizing social media recommender systems with missing recommendations

2025-04-15

AI Summary

This paper addresses the “black-box” challenge in social media recommendation systems—specifically, how to identify implicitly operating recommendation algorithms and quantify their causal impact on societal polarization and misinformation diffusion under conditions of absent ground-truth recommendation logs, opaque system influence, and sparse observational data. Method: We propose the Recommender-Neutral User (RNU) model and a hypothesis-driven synthetic data generation framework (RHSD), which jointly leverage only user interaction network topology and behavioral sequences to enable log-free, end-to-end recommender identification. Our approach integrates graph neural networks, counterfactual modeling, and multi-hypothesis user behavior modeling. Contribution/Results: Evaluated on both synthetic and real-world platform data, our method achieves >92% identification accuracy—substantially outperforming existing auditing techniques—and enables the first interpretable, reproducible, causal attribution analysis of recommendation systems’ societal impacts.

Technology Category

• Data Mining & Knowledge Management: Recommender Systems
• Reasoning under Uncertainty: Causality
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Social media plays a crucial role in shaping society, often amplifying polarization and spreading misinformation. These effects stem from complex dynamics involving user interactions, individual traits, and recommender algorithms driving content selection. Recommender systems, which significantly shape the content users see and decisions they make, offer an opportunity for intervention and regulation. However, assessing their impact is challenging due to algorithmic opacity and limited data availability. To effectively model user decision-making, it is crucial to recognize the recommender system adopted by the platform. This work introduces a method for Automatic Recommender Recognition using Graph Neural Networks (GNNs), based solely on network structure and observed behavior. To infer the hidden recommender, we first train a Recommender Neutral User model (RNU) using a GNN and an adapted hindsight academic network recommender, aiming to reduce reliance on the actual recommender in the data. We then generate several Recommender Hypothesis-specific Synthetic Datasets (RHSD) by combining the RNU with different known recommenders, producing ground truths for testing. Finally, we train Recommender Hypothesis-specific User models (RHU) under various hypotheses and compare each candidate with the original used to generate the RHSD. Our approach enables accurate detection of hidden recommenders and their influence on user behavior. Unlike audit-based methods, it captures system behavior directly, without ad hoc experiments that often fail to reflect real platforms. This study provides insights into how recommenders shape behavior, aiding efforts to reduce polarization and misinformation.

Problem

• Identify hidden social media recommender systems
• Model user behavior influenced by recommenders
• Detect recommender impact on polarization and misinformation

Innovation

• Uses Graph Neural Networks for recommender recognition
• Trains Recommender Neutral User model with GNN
• Generates synthetic datasets for hypothesis testing