Viral video diffusion over the social network

The applet requires Java 5 or higher. Java must be enabled in your browser settings. Mac users must have Mac OS X 10.4 or higher. Windows and Linux users may obtain the latest Java from Oracle's Java site.

WHAT IS IT?

This is a model of diffusion of a potentially viral video through a social network. The social network consist of two sub-networks: the network of friends (undirected small-world network) and the network of followers (directed power network).

HOW IT WORKS

Initially, each agent draws a random number p from a continuous uniform distribution U(0,max-virality). When an agent views the video, she shares it with each of her friends and followers with the probability p. She shares it only once. This way the video diffuses through the network.

HOW TO USE IT

You (as a marketer) estimate the maximal viral potential of the video (max-virality), select a method to seed the agents (initial-seeding: you can select random agents, agents with the maximal number of friends, with the maximal number of followers, the maximal number of friends and followers, or with the maximal expected number of friends and follower the agent shares the video with), and the size of the seed (no-of-initial-infected).

THINGS TO NOTICE

There is phase transition. In phase 1, almost no agents view the video. In phase 2, there is ``epidemic’’—a great proportion of the agents views it.

THINGS TO TRY

Change the structure of the social network, the maximal virality of the video, the seeding strategy, and the seed size, and observe, whether there is an ``epidemic’’, and if so, of what size.

EXTENDING THE MODEL

Not disclosed now. I plan to do it myself. :-)

NETLOGO FEATURES

Power-network generating algorithm. Heavy use of BehaviorSpace.

RELATED MODELS

NetLogo Models Library ``Small Worlds’’ model.

CREDITS AND REFERENCES

Reference will be added after the paper(s) publication.