Which models (that can be represented by means of ODE or discrete counterpart) of social dynamics (behavior, culture, opinion spreading, etc) should absolutely be part of a course on the topic? Network versions only allowed. 🧪🌐🧑‍🤝‍🧑 Which ones are your favorite and a must? Drop a reply please 👇 - ThreadSky

manlius.bsky.social • 36 days ago

Which models (that can be represented by means of ODE or discrete counterpart) of social dynamics (behavior, culture, opinion spreading, etc) should absolutely be part of a course on the topic? Network versions only allowed.
🧪🌐🧑‍🤝‍🧑

Which ones are your favorite and a must?

Drop a reply please 👇

Comments

digthatdata.bsky.social•36 days ago

scan statistics for anomaly detection in dynamic graphs - https://www.cis.jhu.edu/~parky/Enron/

digthatdata.bsky.social•36 days ago

Manifestation of the "you don't date your friends ex" social more on the topology of romantic networks - https://pdodds.w3.uvm.edu/files/papers/others/2002/bearman2002u.pdf

digthatdata.bsky.social•36 days ago

preferential attachment and powerlaw phenomena should make an appearance. Ohh, the friendship paradox is a good one too. https://en.wikipedia.org/wiki/Friendship_paradox

digthatdata.bsky.social•36 days ago

hawkes (self-excitation) process to model content virality https://dl.acm.org/doi/abs/10.1145/3122865.3122874

yyahn.bsky.social•36 days ago

a shameless plug 👀: https://www.science.org/doi/10.1126/sciadv.adh4439

And other belief dynamics models by Mirta et al. And the associative diffusion model by Amir Goldberg

manlius.bsky.social•36 days ago

No shame, many thanks!

pholme.bsky.social•36 days ago

there are more that need to be mentioned than what fits in 300 characters :D But Katz & Lazarsfeld, Granovetter's threshold models, and SIR, feel like a core.

manlius.bsky.social•36 days ago

The real core is already in my hands.. I am looking for something I might have overlooked, and proposals from social scientists are more than welcome.

PA+variations, SIR + variations, voter model + variations are there.

But there are less known yet powerful models out there.. So the survey.

rensec.bsky.social•36 days ago

Perhaps network extensions of Granovetter's threshold models, such as by Watts? And complex contagion (Centola & Macy)...

manlius.bsky.social•36 days ago

Please, reply as a thread

kristinalerman.bsky.social•36 days ago

I am partial to our latest model of opinion dynamics with affective polarization

Simple model, complex results

https://arxiv.org/abs/2403.16940

vitotrianni.bsky.social•35 days ago

Check our best-of-n collective decision model inspired by honeybee nest site selection. Very well studied with ODEs and recently also with heterogeneous mean field modelling by @joefresna.bsky.social

https://journals.aps.org/pre/abstract/10.1103/PhysRevE.95.052411

https://journals.aps.org/pre/abstract/10.1103/PhysRevE.109.054307

praty1931.bsky.social•36 days ago

Edge based compartmental modelling framework for disease spreading on networks. It is an impressive method for reducing a large system of ODEs to a couple of equations.

aespinosarada.bsky.social•36 days ago

In the social science we often learn (longitudinal or temporal) exponential random graph models (ERGMs), stochastic actor-oriented models (SAOMs) and (generalized) relational event models (REM) :)

nwlandry.bsky.social•36 days ago

Bounded confidence models, this paper (https://arxiv.org/abs/2007.00601), and this paper (https://www.nature.com/articles/s41467-019-10431-6)

sandrofsousa.bsky.social•36 days ago

Perhaps Schelling's Model of Segregation? Originally an ABM, but widely studied in a network framework. Schelling, Thomas C. (1971)

manlius.bsky.social•36 days ago

For sure. But is there any ODE-PDE version?
The focus is on modeling dynamics via equations.

sandrofsousa.bsky.social•32 days ago

I've seen some analytical solutions, though, it might not be the most intuitive approach. I was thinking of this work:
https://www.pnas.orgcgidoi10.1073pnas.0609371103

baronca.bsky.social•33 days ago

Two from Axelrod: dissemination of culture, and ethnocentrism. Plus I feel a moral obligation to name the naming game, for which you have models and experiments with humans ;)

manlius.bsky.social•36 days ago

Suggestions?

@nachristakis.bsky.social @lorenzspreen.bsky.social @baronca.bsky.social @sgonzalezbailon.bsky.social @davidlazer.bsky.social @alexvespi.bsky.social @stevenstrogatz.com @carlbergstrom.com @estebanmoro.bsky.social @sunelehmann.com @anxosanchez.bsky.social

anxosanchez.bsky.social•36 days ago

Hi @manlius.bsky.social, I think you can find @psmaldino.bsky.social's book quite useful in this regard. It's a textbook so models are first introduced in well-mixed populations and only afterwards considered on networks but the model selection I find very appropriate

manlius.bsky.social•36 days ago

I missed that, thank you Anxo 🙏

mtizzoni.bsky.social•36 days ago

Following the replies...

lorenzspreen.bsky.social•36 days ago

Spontaneously, I like this overview: https://research.rug.nl/en/publications/models-of-social-influence-towards-the-next-frontiers
maybe some more recent opinion dynamics models:
https://www.tandfonline.com/doi/full/10.1080/0022250X.2018.1517761
https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.124.048301
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0074516
But also discussing complex contagion or models with competition is interesting

leotorresazcona.bsky.social•36 days ago

In ONA, I mostly use static and dynamic models to study collaboration. For diffusion, complex contagion, to understand how behaviors. I’d prioritize SIS/SIR models for diffusion, Threshold models for complex contagion, and Voter models for opinion shifts.

Posting Rules

Be respectful to others
No spam or self-promotion
Stay on topic
Follow Bluesky's terms of service

Comments

Posting Rules

Reply