An individual-based model of information diffusion combining friends' influence
Authored by Lidan Fan, Zaixin Lu, Weili Wu, Yuanjun Bi, Ailian Wang, Bhavani Thuraisingham
Date Published: 2014
DOI: 10.1007/s10878-013-9677-x
Sponsors:
United States National Science Foundation (NSF)
Platforms:
No platforms listed
Model Documentation:
Other Narrative
Pseudocode
Mathematical description
Model Code URLs:
Model code not found
Abstract
In many real-world scenarios, an individual accepts a new piece of
information based on her intrinsic interest as well as friends'
influence. However, in most of the previous works, the factor of
individual's interest does not receive great attention from researchers.
Here, we propose a new model which attaches importance to individual's
interest including friends' influence. We formulate the problem of
maximizing the acceptance of information (MAI) as: launch a seed set of
acceptors to trigger a cascade such that the number of final acceptors
under a time constraint T in a social network is maximized. We then
prove that MAI is NP-hard, and for time , the objective function for
information acceptance is sub-modular when the function for friends'
influence is sub-linear in the number of friends who have accepted the
information (referred to as active friends). Therefore, an approximation
ratio for MAI problem is guaranteed by the greedy algorithm. Moreover, we also prove that when the function for friends' influence is not
sub-linear in the number of active friends, the objective function is
not sub-modular.
Tags
Social networks