Search papers, labs, and topics across Lattice.
This paper introduces TSN4PI, a unified framework designed to track the evolution of political ideologies on social media, addressing challenges like data scarcity and the presence of non-political content. The framework employs a Political Ideology Detection Network (PIDN) that leverages large language models for robust ideology detection and a Political Ideology Prediction Network (PIPN) utilizing temporal graph neural networks for forecasting ideological shifts. Extensive case studies on platforms such as X and Truth Social demonstrate the framework's effectiveness in analyzing political polarization and the dynamics of online ideologies.
Political ideologies on social media can be tracked and predicted with unprecedented accuracy, revealing the complexities of online polarization.
The rapid growth of social media has greatly influenced political discourse, highlighting the need to understand individual political ideologies and their temporal dynamics. This task faces challenges such as data scarcity, abundant non-political content, costly and bias-prone manual annotation, and difficulty in modeling future ideological inclinations. To address these issues, we propose TSN4PI, a unified framework for tracking the evolution of political ideologies on social media. It includes two core modules. The PIDN uses large language models with style transfer and unsupervised domain adaptation to enable robust ideology detection and filter irrelevant content from noisy, cross-domain data. The PIPN employs temporal graph neural networks to predict future ideological shifts, enabling comprehensive analysis of ideology presence, intensity, and evolution. We release two large-scale datasets for noncommercial research use to facilitate further work. Extensive case studies on multiple platforms (X and Truth Social) validate the effectiveness of TSN4PI and provide empirical insights into political polarization and the evolution of online ideologies. Our findings offer a nuanced perspective, advancing both methodological development and empirical understanding in this field.
