Misinformation spread mechanism and control strategies in social media under public emergencies

The pervasive nature of social media has amplified the spread of misinformation, especially under public emergencies. This dissertation investigates the mechanisms underlying the dissemination of misinformation and proposes effective control strategies to mitigate its impact. Through an in-depth analysis of existing literature and case studies, the general propagation patterns of misinformation on social media under public emergencies are examined, particularly through the comparative analysis of the simultaneous propagation of non-misinformation. To better understand the spread mechanism of misinformation on social media under public emergencies, a dynamic model is constructed that represents the spread process and the status of the user. The model includes the debunking process and is validated through real-world social media networks. Additionally, the impacts of different parameters on the spread process are further investigated in simulations. Furthermore, the dissertation proposes strategies to control the spread of misinformation under public emergencies by conducting a comprehensive analysis of users and the propagation path of information, particularly from the task of misinformation detection. A model is built with the integration of user behavior information and information propagation structure, demonstrating improved performance compared to the benchmark misinformation detection models. In general, this dissertation explores the mechanisms of misinformation propagation under public emergencies by examining general propagation patterns, constructing a dynamical model, and proposing strategies for social media platforms to control misinformation through user behavior and information propagation structure analysis.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Informatics

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2024

Author dc:creator

• Yang, Xiujia

Contributors dc:contributor

• Oh, Sang-Hwa
• Yang, Junghwan
• Houston, Brant
• Kilicoglu, Hali

• Social Meida
• Natural Language Processing

Statement dc:rights

• Copyright 2024 Xiujia Yang

Language dc:language

eng, en

Handle dc:identifier

https://hdl.handle.net/2142/127345

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F127345