Misbehaviour Detection Algorithms and Application in Social Networks

Yin, Jun

Misbehaviour Detection Algorithms and Application in Social Networks

Yin, Jun

Permalink

Publication Type:: Thesis
Issue Date:: 2021

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download contents and abstractAdobe PDF (262.82 kB)

Adobe PDF

Download thesisAdobe PDF (6.24 MB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Yin, Jun
dc.date.accessioned	2022-11-03T22:43:01Z
dc.date.available	2022-11-03T22:43:01Z
dc.date.issued	2021
dc.identifier.uri	http://hdl.handle.net/10453/163218
dc.description	University of Technology Sydney. Faculty of Engineering and Information Technology.	en_US.UTF-8
dc.description.abstract	The liberty to contribute content freely has encouraged malicious users to exploit the social platforms (i.e., social networks and e-commerce platforms) for their benefits. Spammers, rumours, and some other unexpected activities are almost an appendage to all social platforms that disrupt the network order. We summarize these unexpected activities as misbehaviours in social platforms. To detect such social platform misbehaviours, machine learning is an expected method where modelling and algorithms are two significant elements. Such an interesting topic that has application prospects and research value has attracted the attention of many researchers, and some results have also been put forward in the literature. In terms of spammer detection, due to the special characteristic, social networks call for relation-dependent but content-independent methods for spammer detection. To make up for the lack of existing research, this thesis proposes two methods for spammer detection on multi-relational social networks. First, “Send-Receive” Role Separable Graph-Embedding Model (RS-GEM) to extract and fuse the hidden features of heterogeneous relations in multi-relational social networks. Second, Multi-level Dependency Model (MDM), which exploits user’s behaviours in terms of long-term and short-term dependency from both individual-level and union-level. As for rumour analysis, before a rumour has an impact on social networks, we need to assess the possible impact it may have. Therefore, we devise a rumour influence prediction model RISM (Rumour Impact on Social Media) based on a popular rumour intensity formula to predict the impact of a newborn rumour. At last, since the global outbreak of COVID-19 in early 2019, COVID-19-related topics have become hot spots on social networking platforms. We analyze the COVID-19-related tweets on Twitter and get a preliminary understanding of the public's focus and sentiment trends during the pandemic.	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/10453/163218/2/02whole.pdf
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	au.edu.uts.lib/ppc
dc.rights	info:eu-repo/semantics/openAccess
dc.title	Misbehaviour Detection Algorithms and Application in Social Networks	en_US.UTF-8
dc.type	Thesis
utslib.copyright.status	open_access	*

Abstract:

The liberty to contribute content freely has encouraged malicious users to exploit the social platforms (i.e., social networks and e-commerce platforms) for their benefits. Spammers, rumours, and some other unexpected activities are almost an appendage to all social platforms that disrupt the network order. We summarize these unexpected activities as misbehaviours in social platforms. To detect such social platform misbehaviours, machine learning is an expected method where modelling and algorithms are two significant elements. Such an interesting topic that has application prospects and research value has attracted the attention of many researchers, and some results have also been put forward in the literature. In terms of spammer detection, due to the special characteristic, social networks call for relation-dependent but content-independent methods for spammer detection. To make up for the lack of existing research, this thesis proposes two methods for spammer detection on multi-relational social networks. First, “Send-Receive” Role Separable Graph-Embedding Model (RS-GEM) to extract and fuse the hidden features of heterogeneous relations in multi-relational social networks. Second, Multi-level Dependency Model (MDM), which exploits user’s behaviours in terms of long-term and short-term dependency from both individual-level and union-level. As for rumour analysis, before a rumour has an impact on social networks, we need to assess the possible impact it may have. Therefore, we devise a rumour influence prediction model RISM (Rumour Impact on Social Media) based on a popular rumour intensity formula to predict the impact of a newborn rumour. At last, since the global outbreak of COVID-19 in early 2019, COVID-19-related topics have become hot spots on social networking platforms. We analyze the COVID-19-related tweets on Twitter and get a preliminary understanding of the public's focus and sentiment trends during the pandemic.

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

http://hdl.handle.net/10453/163218