On the Use of Network Control Techniques in Pursuit of Influence Spread in Complex Networks

Sadaf, Abida

On the Use of Network Control Techniques in Pursuit of Influence Spread in Complex Networks

Sadaf, Abida

Permalink

Publication Type:: Thesis
Issue Date:: 2024

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download thesisAdobe PDF (3.55 MB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Sadaf, Abida
dc.date.accessioned	2024-04-16T01:09:18Z
dc.date.available	2024-04-16T01:09:18Z
dc.date.issued	2024
dc.identifier.uri	http://hdl.handle.net/10453/177935
dc.description	University of Technology Sydney. Faculty of Engineering and Information Technology.	en_US.UTF-8
dc.description.abstract	A critical element in the process of control and influence spread is the selection of the seed nodes from which influence and control spreads. To be able to meet project aim and to develop more effective seed selection methods, first we need to understand the relationship between different global and local network structures and the number of driver nodes needed to control a given structure. This reveals what structures are easier to control and the resources needed to control them. The first component of the thesis highlights how differing structure in both real and synthetic social networks affects the number of driver nodes needed for control. We investigate a correlation between global structural measures and the number of driver nodes. Experiments show that there is a strong relationship between density and the number of driver nodes. Next, the thesis investigates how the number of driver nodes identified at the level of individual communities relates to the densities of those communities. This illustrates how local structures and their composition influence the number of driver nodes. . Using this base set, we propose new methods based (i.e. Driver-Random, Driver-Degree, Driver-Closeness, Driver-Betweenness, Driver-Degree-Closeness-Betweenness, Driver-Kempe, Driver-Ranked) for selecting seeds. These methods make use of network centrality measures to rank the driver nodes in terms of their potential as seed nodes. As a result we get a small subset of driver nodes, that helps in improving influence spread. the final study uses ‘divide and conquer’ approach to the time-consuming problem of driver node identification at the global level and instead identifies driver nodes within the communities, then using those driver nodes in the influence spread process. In this thesis we demonstrate the effectiveness of this approach in Random, Small-World and Scale-Free networks as well as real-world social networks. The process begins with identification of communities within the network, and then identification of driver nodes for each community separately. The driver nodes obtained are then ranked according to a range of common centrality measures (using similar protocol as with the whole network approach). We then compare the total number of nodes influenced as a result of utilising various seed selection methods based upon globally elected and ranked driver nodes and locally selected and ranked driver nodes. This approach is not only novel in its basic concept, but provides improved algorithmic outcomes alongside more effective influence spread.	en_US.UTF-8
dc.format	Thesis (PhD)
dc.language.iso	en_US	en_US.UTF-8
dc.relation	https://opus.lib.uts.edu.au/bitstream/10453/177935/1/thesis.pdf
dc.rights	info:eu-repo/semantics/openAccess
dc.rights	The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.
dc.rights	© 2024 Abida Sadaf
dc.rights	au.edu.uts.lib/nph
dc.title	On the Use of Network Control Techniques in Pursuit of Influence Spread in Complex Networks	en_US.UTF-8
dc.type	Thesis
utslib.copyright.status	open_access	*

Abstract:

A critical element in the process of control and influence spread is the selection of the seed nodes from which influence and control spreads. To be able to meet project aim and to develop more effective seed selection methods, first we need to understand the relationship between different global and local network structures and the number of driver nodes needed to control a given structure. This reveals what structures are easier to control and the resources needed to control them. The first component of the thesis highlights how differing structure in both real and synthetic social networks affects the number of driver nodes needed for control. We investigate a correlation between global structural measures and the number of driver nodes. Experiments show that there is a strong relationship between density and the number of driver nodes. Next, the thesis investigates how the number of driver nodes identified at the level of individual communities relates to the densities of those communities. This illustrates how local structures and their composition influence the number of driver nodes. . Using this base set, we propose new methods based (i.e. Driver-Random, Driver-Degree, Driver-Closeness, Driver-Betweenness, Driver-Degree-Closeness-Betweenness, Driver-Kempe, Driver-Ranked) for selecting seeds. These methods make use of network centrality measures to rank the driver nodes in terms of their potential as seed nodes. As a result we get a small subset of driver nodes, that helps in improving influence spread. the final study uses ‘divide and conquer’ approach to the time-consuming problem of driver node identification at the global level and instead identifies driver nodes within the communities, then using those driver nodes in the influence spread process. In this thesis we demonstrate the effectiveness of this approach in Random, Small-World and Scale-Free networks as well as real-world social networks. The process begins with identification of communities within the network, and then identification of driver nodes for each community separately. The driver nodes obtained are then ranked according to a range of common centrality measures (using similar protocol as with the whole network approach). We then compare the total number of nodes influenced as a result of utilising various seed selection methods based upon globally elected and ranked driver nodes and locally selected and ranked driver nodes. This approach is not only novel in its basic concept, but provides improved algorithmic outcomes alongside more effective influence spread.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/177935