Computational Model Library

Linear Threshold (1.0.0)

WHAT IS IT?

This is an agent based implementation of the Linear Threhsold model of influence propagation described in Kempe et al. (2003).

HOW IT WORKS

The influence propagation process starts with a social network and an initial set of active agents (seeds or early adopters), and then at each time step those active ageents try to influence other agents into converting into active adopters according to the influence weights of their link. Depending on the threshold of an agent it may or may not become active. See the attached ODD for more information.

HOW TO USE IT

Fist set the number of nodes and number of initial seeds. Then choose the type of network to experiment with. If the chosen network type is “Small World” of “Spatially Clustered”, then you can further change the generation parameters. After the network generation parameters are properly entered, click on the “setup network” button to create the network in the screen. Next you can either choose the initial set of adopters (seeds) randomly by clicking the the “select seeds randomly” button, or click the “select seeds by mouse” button and use the mouse to select seeds. Selecting seeds in this way will override the number-of-seeds parameter (i.e. you can choose more number of agents or less number of agents than the number-of-seeds parameter). Next hit “go” to run the model. To resent the thresholds for a new run click the “reset” button. By turning the “threshold-fixed-to-0.5” button you can configure the model to assign a fixed threshold of 0.5 to each agent. If this switch is turned off then each time reset will assign random values between 0 and 1.00 to assign to the threshold values of the agent.

THINGS TO NOTICE

See how the final number of active agents change over different runs. Also notice how the number of active users change over time (ticks elapsed).

EXTENDING THE MODEL

Add more options for creating the network. One new alternative is to read an external network written in graphML, pajek or SNAP format from a file.

“Virus on a Network” model in NetLogo Model Library.

CREDITS AND REFERENCES

The original model was described in the following paper:

Maximizing the spread of influence through a social network - Kempe D, Kleinberg J,
Tardos E, SIGKDD 2003

Replicated by Kaushik Sarkar.

Release Notes

Associated Publications

Linear Threshold 1.0.0

WHAT IS IT?

This is an agent based implementation of the Linear Threhsold model of influence propagation described in Kempe et al. (2003).

HOW IT WORKS

The influence propagation process starts with a social network and an initial set of active agents (seeds or early adopters), and then at each time step those active ageents try to influence other agents into converting into active adopters according to the influence weights of their link. Depending on the threshold of an agent it may or may not become active. See the attached ODD for more information.

HOW TO USE IT

Fist set the number of nodes and number of initial seeds. Then choose the type of network to experiment with. If the chosen network type is “Small World” of “Spatially Clustered”, then you can further change the generation parameters. After the network generation parameters are properly entered, click on the “setup network” button to create the network in the screen. Next you can either choose the initial set of adopters (seeds) randomly by clicking the the “select seeds randomly” button, or click the “select seeds by mouse” button and use the mouse to select seeds. Selecting seeds in this way will override the number-of-seeds parameter (i.e. you can choose more number of agents or less number of agents than the number-of-seeds parameter). Next hit “go” to run the model. To resent the thresholds for a new run click the “reset” button. By turning the “threshold-fixed-to-0.5” button you can configure the model to assign a fixed threshold of 0.5 to each agent. If this switch is turned off then each time reset will assign random values between 0 and 1.00 to assign to the threshold values of the agent.

THINGS TO NOTICE

See how the final number of active agents change over different runs. Also notice how the number of active users change over time (ticks elapsed).

EXTENDING THE MODEL

Add more options for creating the network. One new alternative is to read an external network written in graphML, pajek or SNAP format from a file.

“Virus on a Network” model in NetLogo Model Library.

CREDITS AND REFERENCES

The original model was described in the following paper:

Maximizing the spread of influence through a social network - Kempe D, Kleinberg J,
Tardos E, SIGKDD 2003

Replicated by Kaushik Sarkar.

Version Submitter First published Last modified Status
1.0.0 Kaushik Sarkar Sat Nov 3 06:57:06 2012 Fri Feb 16 22:23:34 2018 Published

Discussion

This website uses cookies and Google Analytics to help us track user engagement and improve our site. If you'd like to know more information about what data we collect and why, please see our data privacy policy. If you continue to use this site, you consent to our use of cookies.
Accept