Seldon: Opinion dynamics toolkit

Understanding the dynamics of opinions within social networks is of paramount importance in various fields, from sociology to political science and beyond. We present a versatile and efficient C++ code for modeling opinion dynamics that not only encapsulates an array of well-established empirical models but also offers powerful post-processing and plotting scripts. Interestingly, analogous systems have been studied in various, well established, branches of physics.

We present Seldon [1], an open source, high performance computing friendly C++ engine and Python plotting toolkit for melding insights from disparate domains (computer science, machine learning, humanities) into a cohesive package. Our design encompasses classic models like the De Groot [2] model, the Voter [3] model, and more recent activity driven models [4], while being fully extensible. The study of phase transitions in such systems will also be briefly discussed in relation to the classic Ising model.

Refs:

[1] https://github.com/seldon-code/

[2] M.H. DeGroot, Reaching a consensus, J. Am. Stat. Assoc., 69 (1974)

[3] Clifford et. al, A Model for Spatial Conflict. Biometrika 60 (1973)

[4] Baumann et. al, Modeling Echo Chambers and Polarization Dynamics in Social Networks, Phys. Rev. Lett. 124 (2020)