samthiriot / model-wom-usa-ipk

a model of word-of-mouth

What is it?

This is the implementation of a model of word-of-mouth. In this model, we investigate what happens is we do not only represent information being just diffused amongst a population, but when individuals also search for information.

The model is described in the following publication:

Samuel Thiriot, Word-of-mouth dynamics with information seeking: Information is not (only) epidemics, Physica A: Statistical Mechanics and its Applications, Volume 492, 2018, Pages 418-430, ISSN 0378-4371, https://doi.org/10.1016/j.physa.2017.09.056. (http://www.sciencedirect.com/science/article/pii/S0378437117309482)

How to use it?

This model runs under Netlogo 6; you can freely download it from here: https://ccl.northwestern.edu/netlogo/download.shtml

Start Netlogo; open the model-wom-information-seeking.nlogo file from netlogo; click setup, then simulate. Then restart by tuning parameters. Have a look to the "info tab" which explains what happens and what to explore.

Exploration of the space of parameters

To explore the space of parameters, we rely on the OpenMole software: https://www.openmole.org/

Open in OpenMole the exploration workflow found in the "exploration" directory. For each point of the space of parameters, this workflow will:

• generate a network using R/igraph and measure its properties
• transmit it to netlogo as an input file
• store the simulation results into a CSV file

Advanced

generate other networks

networks in this model represent the structure of interactions, which has a huge influence on the dynamics simulated.

You can easily genarate novel graphs to use using R/igraph.

R
library(igraph)
g <- igraph::watts.strogatz.game(size=1000,nei=4,dim=1,multiple=F,loops=F,p=0.1)
average.path.length(g)
write.graph(g, file="network_1000.graphml", format="graphml")