Information Technology and the Propagation of Racism in the Australian Community


Supervisory Chair



This project seeks to apply and/or advance the state of the art in computer science in the detection and prevention of use of information technology for racist purposes. The project may involve basic research, applied research or a combination of both. It is likely that existing methods addressing other online threats will be relevant to the problem domain including methods to address online misogyny, bullying, terrorist networks and their identification and disruption (as examples). The project is likely to involve use of natural language processing, machine learning and network analysis as existing technologies deployed in this problem domain. 


The project may involve working with government and non-government external stakeholders concerned with the problem of online racism to assist them in responding to online racism.  Research by publication is encouraged as part of this project.


Some Background


In November 2018 the founder of the world wide web, Tim Berners Lee launched a campaign to "save the web". The campaign called on institutional and individual users of the web to sign on to a "Contract for the Web".  The background of the launch was the misuse of the web for purposes never intended by its founders. Among these is the propagation of racism (including online racist bullying), which is typically a violation of the terms of use associated with the internet and services provided via the internet. Among the principles of the contract for the web were an undertaking by companies to:


  • Develop technologies that support the best in humanity and challenge the worst
  • So the web really is a public good that puts people first.
And an undertaking by individuals to Build strong communities that respect civil discourse and human dignity So that everyone feels safe and welcome online.
The propagation of racism through the internet and other information technologies at its worst represents a threat to democratic institutions. It directly impacts and harms individuals and communities targeted by online racism and often harms the well-being of those who are recruited into racist networks. Political movements seeking to establish or re-establish preferential treatment on the basis of race or exploit racist sentiment for political purposes use information technology to propagate their messages. Such movements have demonstrated their ability to affect overall societal trends. Companies such as Facebook and Twitter have been criticised for their inadequate response to the use of their platforms for racist purposes and in other ways which bully or harass individuals and communities. 
Within Australia, racist messages take on a particular character, sometimes directed against indigenous Australians (who are among those most impacted by online racism) and recent migrants. Debates concerning contested social issues such as immigration or crime are exploited to promote racist ideas and attitudes. Sometimes the Australian context consciously or unconsciously absorbs racist ideas from broader global discourse. Some racist discourse is generated by media coverage and political figures. In 2018 racist networks were alleged to have infiltrated a major Australian political party (the Nationals). The networks were identified by investigative journalism  
Outline of Some Potential Research Issues
The detection of and response to racism in online environments is a growing field of research. Beyond detection and description of racism is the development of effective mechanisms to impede the spread of racism in online fora. Information technology can also be used in a positive sense. Racist discourse is of course generated by human beings. Development of effective and objective measures for assessing whether particular speech is racist (or has racist impacts) is a philosophical, linguistic and computational problem that would benefit from research and attention. Any appropriate addressing of such issues must be based on an understanding of the boundaries of appropriate free speech in a democracy and how appropriate speech can be distinguished from speech which directly or indirectly has a racist content or impact. Inevitably research in this area must also begin from a theory of the nature and character of racism.  As an example of problems in the domain, research in 2018 showed the fragility of existing machine learning methods to simple strategies for evading detection. Researchers involved also noted the inherent subjectivity of training datasets for what constitutes "hate speech". ( Detection moreover is of limited value if tools do not exist to respond to disrupt/interdict online harassment. A key problem in that respect is the identification of online abusers who are motivated to seek to evade detection (e.g. taking advantage of the anonymity of the web). 
Examples of existing research that relates to this project are linked below.
We are seeking expressions of interest from students who may be interested in undertaking research at the Masters or PhD level, related to the issues outlined above.  Students would be supported by a multidisciplinary team of supervisors. 


Background Literature

Hate Speech Detection Using Natural Language Processing Techniques


Requiem for online harassers: Identifying racism from political tweets
All You Need is “Love”: Evading Hate Speech Detection

Multilingual Cross-domain Perspectives on Online Hate Speech





Mining communities and their relationships in blogs: A study of online hate groups
Blogs, often treated as the equivalence of online personal diaries, have become one of the fastest growing types of Web-based media. Everyone is free …




Platformed racism: The mediation and circulation of an Australian race-based controversy on Twitter, Facebook and YouTube





Hateful Symbols or Hateful People? Predictive Features for Hate Speech Detection on Twitter

Proceedings of NAACL-HLT 2016, pages 88–93, San Diego, California, June 12-17, 2016. c 2016 Association for Computational Linguistics Hateful Symbols or Hateful People?
Matamoros-Fernandez, Ariadna < (2016) Platformed racism: The mediation and circulation of an Australian race-based controversy on Twitter, Facebook and YouTube. In Association of Internet Researchers Annual Conference (AoIR 2016), 5-8 October 2016, Berlin. (Unpublished)

Modeling Community Behavior through Semantic Analysis of Social Data: the Italian Hate Map Experience
Practical Algorithms for Destabilizing Terrorist Networks

A Unified Deep Learning Architecture for Abuse Detection






Deep Learning for Hate Speech Detection in Tweets
Hate speech detection on Twitter is critical for applications like controversial event extraction, building AI chatterbots, content recommendation, and sentiment analysis.



Cyberbullying Detection and Classification Using Information Retrieval Algorithm
Social networking site is being rapidly increased in recent years, which provides platform to connect people all over the world and share their interests.




Social media and digital technology use among Indigenous young people in Australia: a literature review
The use of social media and digital technologies has grown rapidly in Australia and around the world, including among Indigenous young people who face social disadvantage. Given the potential to use social media for communication, providing information and as part of creating and responding to ...




Issues Surrounding Cyber-Safety for Indigenous Australians
Joint Select Committee on Cyber-Safety


Content Analysis KIMBERLY A. NEUENDORF and ANUP KUMAR Cleveland State University, USA Introduction ...
Alt_Right White Lite: trolling, hate speech and cyber racism on social media
2017] Regulating Cyber-Racism 3 Advance Copy deluge of online animosity that flooded their Facebook page.1 Goodes, a retired Australian Football League (‘AFL’) player, had already been the target
Racist violence in schools is on the documented increase worldwide. This paper will make the argument that the nature of and motivations for such attacks are changing as a function of the new electronic communication technologies available to students. The prevalence in school communities is thought to be under-represented due to the under-reporting of incidents to authorities.

Islamophobia and Twitter: A Typology of Online Hate Against Muslims on Social Media
Cyber bullying, cyber stalking, human rights violations online and related domains

Updated:  1 June 2019/Responsible Officer:  Dean, CECS/Page Contact:  CECS Marketing