Infodemic Source Detection with Information Flow: Foundations and Scalable Computation

Area of Science:

• Network Science
• Information Theory
• Computational Social Science

Background:

• Traditional rumor source identification methods, like maximum likelihood (ML) and joint maximum likelihood (JML) estimators using the Susceptible-Infectious (SI) model, suffer from degeneracy.
• These classical approaches often fail to uniquely identify the rumor source, even in basic network configurations.

Purpose of the Study:

• To develop a more robust and accurate method for identifying the source of rumors in networks.
• To overcome the limitations of existing estimators by incorporating random observation times and advanced network flow concepts.

Main Methods:

• Proposed a generalized estimator incorporating independent random observation times.
• Formulated information flow beyond simple graphs, considering rate constraints and multicast capacities for cyclic polylinking networks.
• Developed forward elimination and backward search algorithms for rate-constrained source detection.

Main Results:

• The generalized estimator demonstrates improved performance in identifying rumor sources compared to classical methods.
• Simulations validate the effectiveness and scalability of the proposed algorithms for rate-constrained source detection.
• The study provides a rigorous foundation for infodemic source detection in complex network environments.

Conclusions:

• The novel generalized estimator offers a significant advancement in rumor source detection, particularly in challenging network structures.
• The developed algorithms are effective and scalable, providing practical tools for analyzing information spread.
• This research establishes a robust framework for understanding and mitigating the impact of infodemics.