SocialSim

Social Simulation for Evaluating Online Messaging Campaigns

Last updated on Sep 26, 2022
Credit: By GlowArtbox on smarterpix.com.

Synopsis

Bad faith actors routinely leverage the global information environment to further their agendas by manipulating public opinion, for example, via disinformation. Being able to accurately simulate the spread and evolution of information cascading through the online information domain, from the bottom-up and at-scale, would enhance our fundamental undestanding of the nature of the information diffusion phenomenon that emerges. Such understanding could facilitate, for example, the detection, analysis and countering of strategic disinformation campaigns, as well as determining effective approaches in broadcasting critical disaster relief information to local populations in need. DARPA’s SocialSim program aimed at making breakthroughs along this direction.

Our MLRG joined a coalition of institutions, namely, University of Central Florida, University of North Carolina Charlotte, North Carolina State University, University of New York and Northern Illinois University, and participated in SocialSim as a TA1 (simulation) performer.

Over the course of the project, the team pioneered two simulation models, namely the Multi-Action Cascade Model (MACM) and the Multiplexity-Based Model (MBM), which were refined over time to address six challenge problems involving a variety of scenarios and multiple social media platforms. These models demonstrated plausible social dynamics and were among the well-preforming approaches of the SocialSim program.

People

The overall project’s investigating team consisted of its

Both Xi Zhang and Akshay Aravamudan from our MLRG were involved in various capacities during the course of the project.

Support

Our MLRG’s SocialSim-related efforts were supported by the following grant:

  • Anagnostopoulos (Institutional PI), $432,459, Defense Advanced Research Projects Agency (subcontract from UCF), Prime (Federal) Grant No. FA8650-18-C-7823, “Deep Agent: A Framework for Information Spread and Evolution in Social Networks,” 10/10/2017 – 07/31/2021.

The project supported Zhang over its three-and-a-half year course.

Outcomes

Overall our group produced or contributed to one poster presentation and four conference papers related to the SocialSim effort. It is worth mentioning that the work presented in (zhang2020) was selected for inclusion into a DARPA Information Innovation Office (I2O) briefing presented to the US Undersecretary of Defense for Research and Engineering in August in 2020.

Finally, our involvement into SocialSim was a catalyst in us acquiring substantial expertise in modeling event data using temporal point processes.

Georgios C. Anagnostopoulos
Georgios C. Anagnostopoulos
Associate Professor of Electrical & Computer Engineering

I lead the Machine Learning Research Group at FIT.

Publications

Anytime Information Cascade Popularity Prediction via Self-Exciting Processes
Based on newly derived expressions of conditional mean counts of marked Hawkes point processes, we introduce Cascade Anytime Size Prediction via self-Exciting Regression (CASPER) for anytime information cascade popularity prediction.
PDF Cite Code Project Poster Slides Video
Deep Agent: Studying the Dynamics of Information Spread and Evolution in Social Networks
This work delineates the main concepts and models utilized in our social media modeling effort under DARPA’s SocialSim program.
Cite Project
Influence Dynamics Among Narratives
In this work we employ multi-variate Hawkes processes and introduce process influence measures to analyze the interactions of competing and coevolving Twitter narratives surrounding the 2019 Venezuelan Presidential Crisis.
Cite Project Slides Video DOI
Predicting Software Vulnerability Exploits from Social Media Confabulations
In this work we share some of our preliminary findings about predicting the timing of software vulnerability exploitations based on the vulnerabilities’ social media popularity.
PDF Cite Project Poster Video
Deep Agent: Computational Social Science Centered Simulation of Online Information Environments
In this paper we lay out the design of a social network analysis framework, developed under DARPA’s SocialSim program. Our novel architecture models human emotional, cognitive and other social factors. Furthermore, we showcase results of modeling cryptocurrencies, cyber threats, and software vulnerabilities for three interrelated social environments, namely GitHub, Reddit, and Twitter.
PDF Cite Project