Print

Print


Hi Ali,

If you want code samples, your best bet is to take a look at the epidemic models that come with NetLogo. You can then translate those into MASON fairly straightforwardly if you know Java (or even scala if you're feeling adventuresome)

They all pretty much work like what John has said. I have a model that uses gossip algorithms to solve the distributed consensus problem, but like him I'm in the process of using it to publish before I make it public. 

John also makes a very good point about being careful as to when you actually update the agents. If you update them as they're exposed to information, you can get totally different results compared to buffering the updates until the end of the step, then having everybody update "in parallel"


Good luck!


On Thu, Mar 7, 2013 at 5:38 PM, John McManus <[log in to unmask]> wrote:

I need to publish before sending out the code. However, the basic idea is that each node is an agent. Each agent is tied to some proportion of the others randomly using the links in Mason. When that agent is stepped, it polls all the other agents linked to it to count up how many have been innovated already. If that count (or percent of ‘friends’) equals or exceeds the minimum for adoption, that agent adopts the innovation. The trick is to be sure the adoptions are only implemented at the end of a simulation step, lest the field change within a single simulation step. Thus, we mark the agent as “ready to adopt”, but they all actually change state in a loop at the end of the simulation step.  Of course, we start the simulation with a certain number of agents already implemented, using leaders, random or fringe agents (based sorting by how many links each agent has).

 

Good luck!

 

John

 

 

From: MASON Multiagent Simulation Toolkit [mailto:[log in to unmask]] On Behalf Of Ali Nazemian
Sent: Thursday, March 07, 2013 3:35 PM
To: [log in to unmask]
Subject: Re: Simulating spreading the influence in social network

 

Dear John, Thank you for your reply. Actually i am aware of the diffusion models and different types of them. I am looking for a sample code which could accelerate my implementation. It could help me a lot if something like that be available.

Thank you very much.

 

On Thu, Mar 7, 2013 at 10:53 PM, John McManus <[log in to unmask]> wrote:

I have done something like this using Mason, which can help one set up simple social network links and diagrams.

 

For a simple introduction to turning social networks into diffusion models, see:

Valente, T. W., & Davis, R. L. (1999). Accelerating the diffusion of innovations using opinion leaders.  The Annals of the American Academy of the Political and Social Sciences, 566, 55-67.

 

Tom Valente’s recent book is also recommended:

Valente TW (2010) Social Networks and Health: Models, Methods, and Applications [Hardcover], Oxford University Press

 

 

Cheers!

 

 

John

 

John W. McManus, PhD

Director, National Center for Coral Reef Research (NCORE)

Professor, Marine Biology and Fisheries

Coral Reef Ecology and Management Lab (CREM Lab)

Rosenstiel School of Marine and Atmospheric Science (RSMAS)

University of Miami, 4600 Rickenbacker Causeway, Miami, 33149

[log in to unmask]      http://ncore.rsmas.miami.edu/

Phone: 305-421-4814  

 

"Far better an approximate answer to the right question, which is often vague,

   than an exact answer to the wrong question, which can always be made precise."

             

     --John Tukey, Statistician, National Medal of Science and IEEE Medal of Honor

 

 

 

 

From: MASON Multiagent Simulation Toolkit [mailto:[log in to unmask]] On Behalf Of Ali Nazemian
Sent: Thursday, March 07, 2013 5:37 AM
To: [log in to unmask]
Subject: Simulating spreading the influence in social network

 

Hi,

I am looking for a tool to simulating the spread of influence and behaviors in social network, something like viral marketing. Would you please tell me that such a simulation is possible with using MASON? is there any sample project that could be help me in this case?

Regards.

 

--
A.Nazemian



 

--
A.Nazemian