Sequential seeding in multilayer networks

TL;DR

This paper introduces and evaluates a sequential seeding method for multilayer networks, demonstrating it improves influence spread coverage and efficiency over traditional methods, despite increasing process duration.

Contribution

It develops and assesses a novel sequential seeding technique specifically designed for multilayer networks, extending prior work limited to single-layer networks.

Findings

Sequential seeding increases influence coverage in multilayer networks.

The method saves seeding budget compared to traditional approaches.

It extends the duration of the spreading process.

Abstract

Multilayer networks are the underlying structures of multiple real-world systems where we have more than one type of interaction/relation between nodes: social, biological, computer, or communication, to name only a few. In many cases, they are helpful in modelling processes that happen on top of them, which leads to gaining more knowledge about these phenomena. One example of such a process is the spread of influence. Here, the members of a social system spread the influence across the network by contacting each other, sharing opinions or ideas, or - explicitly - by persuasion. Due to the importance of this process, researchers investigate which members of a social network should be chosen as initiators of influence spread to maximise the effect. In this work, we follow this direction, develop and evaluate the sequential seeding technique for multilayer networks. Until now, such techniques were evaluated only using simple one layer networks. The results show that sequential seeding in multilayer networks outperforms the traditional approach by increasing the coverage and allowing to save the seeding budget. However, it also extends the duration of the spreading process.