Throughout my career, I’ve been amazed with the productivity of interdisciplinary teams. Subject matter experts look out of their silos to cooperatively provide unique solutions to difficult problems. Thus, I was intrigued with a paper entitled, “Generalized gelation theory describes onset of online extremist support” (Manrigue, P.D.; Zheng, M. et al. Phys. Rev. Lett. July 27, 2018, 121, 048301). Three of the authors are in the Physics Departments at the University of Miami (FL) and another is from George Washington University (GWU, Washington, DC). The remaining two are from the Elliott School of International Affairs at GWU. This is clearly a heterogeneous bunch.
The models are somewhat different. Gelation can involve homophily only since the gel forms from a homogeneous mixture. With extremist groups, high homophily clearly dominates the most influential groups as measured by the number of follows on VKontakte.
Interestingly, the gelation theory compares well with the growth of group size for extremist support. An analogy between the growth of online networks and the formation of gels suggests ways to detect and predict growth of online extremist groups before they become influential.
VKontakte records the online ISIS activity of different sites in the study time period, which is January through February 2015. During this period, ISIS support grew from 40,000 to 70,000 internode communications (Figure 1). In short, homogeneous content communicated to a large hemophilic base is a predictor of strength. Low activity including following begets low impact and importance, as shown in Figure 2.
Figure 1 – Pro-ISIS communication patterns recorded by VKontakte on January 10, 2015. Intense communication links indicates aggregation of interest.
Figure 2 – Gelation theory correlates with growth and size of online pro-ISIS groups. Circles are the number of follows recorded on VKontakte. Dark lines are predicted behavior from generalized gelation theory that accounts for heterogeneous nature due to humans in support groups. Rapid growth (high slope and number of follows) indicates high homophily and gelation. Groups with low # follows are splinter cells that are not popular despite high activity.I’m impressed that a model that is used to describe aggregation of gels also has utility in predicting the most influential human groups. This tool may help focus on which groups in the fragmented extremist segment merit attention. The authors claim that the gelation model is more accurate in predicating group success than standard aggregation theory.
Interdisciplinary teams usually have closer associations than physical chemistry of gelation and international affairs such as growing support of ISIS in the focus period. Neil F. Johnson is the last listed author of the paper. Johnson joined the faculty at GWU in 2018 from the University of Miami, which explains the affiliations. According to Wikipedia, Johnson is leading a “new initiative in Complexity and Data Science which combines cross-disciplinary fundamental research with data science, with a view to resolving complex real-world problems.” Gelation and ISIS are real-world problems.
Robert L. Stevenson, Ph.D., is Editor Emeritus, American Laboratory/Labcompare; e-mail: [email protected]