Effectiveness of Diffusing Information through a Social Network in Multiple Phases

TL;DR

This paper investigates how multiple phases of seed node selection in social networks can enhance information spread, revealing that two to three phases often optimize diffusion with nearly equal budget splits, though more phases do not always guarantee better results.

Contribution

It provides a detailed simulation analysis of multiphase diffusion, highlighting the practical benefits and optimal budget splitting strategies in real-world networks.

Findings

Two to three phases yield significant diffusion gains.

Equal budget splitting across phases is near-optimal.

More than three phases offers diminishing returns.

Abstract

We study the effectiveness of using multiple phases for maximizing the extent of information diffusion through a social network, and present insights while considering various aspects. In particular, we focus on the independent cascade model with the possibility of adaptively selecting seed nodes in multiple phases based on the observed diffusion in preceding phases, and conduct a detailed simulation study on real-world network datasets and various values of seeding budgets. We first present a negative result that more phases do not guarantee a better spread, however the adaptability advantage of more phases generally leads to a better spread in practice, as observed on real-world datasets. We study how diffusing in multiple phases affects the mean and standard deviation of the distribution representing the extent of diffusion. We then study how the number of phases impacts the effectiveness of multiphase diffusion, how the diffusion progresses phase-by-phase, and what is an optimal way to split the total seeding budget across phases. Our experiments suggest a significant gain when we move from single phase to two phases, and an appreciable gain when we further move to three phases, but the marginal gain thereafter is usually not very significant. Our main conclusion is that, given the number of phases, an optimal way to split the budget across phases is such that the number of nodes influenced in each phase is almost the same.