Detecting Child Sexual Abuse Material Online

The online distribution of child sexual abuse videos and images (also known as child sexual exploitation and child pornography) increases each year. Combined with the graphic nature of the material, the amount of media that needs to be reviewed by investigators outpaces human resources. Over the past decade, there has been an increase in automated tools to identify child sexual abuse material (CSAM). These tools typically rely on keywords and hash values. Unfortunately, both of these are severely limited. Specifically, their inflexibility to change, inability to detect new CSAM, and inability to link CSAM victims and/or offenders means that other techniques need to be prioritized. 

Our research focuses on the automation of searching, identifying, collecting, analyzing, and matching CSAM on both the Surface and Deep/Dark Web. This began with the development of an automated tool to collect potential CSAM following the evolution of CSAM networks, applying social network analysis techniques to identify key players and network traits, identifying website attributes common on CSAM websites, and most recently using biometric attributes (face and voice).

RESEARCH PARTNERS

Associate Professor Russell Brewer, University of Adelaide
Mr Thomas Swearingen, Post-Doctoral Fellow, University of Adelaide
Professor Arun Ross, Michigan State University
Lecturer Katie Lagos, University of Adelaide
Assistant Professor George Azzopardi, University of Groningen
Associate Professor Richard Frank, Simon Fraser University
Professor David Bright, Deakin University
Dr Dana Michalski, Defence Science and Technology Group
Bertram Lyons and Brandon Epstein, Medex Forensics

Publications

  • Logos, K., Brewer, R., Langos, C., & Westlake, B. (2023). Establishing a framework for the ethical and legal use of web scrapers by cybercrime and cybersecurity researchers: Learnings from a systematic review of Australian research. International Journal of Law and Information Technology, eaad023. https://doi.org/10.1093/ijlit/eaad023 [PDF]
  • Westlake, B., & Guerra, E. (2023). Using file and folder naming and structuring to improve automated detection of child sexual abuse images on the Dark Web. Forensic Science International: Digital Investigation, 47(December), 301620. https://doi.org/10.1016/j.fsidi.2023.301620 [PDF]
  • Brewer, R., Westlake, B., Swearingen, T., Patterson, S., Bright, D., Ross, A., Logos, K., & Michalski, D. (2023). Advancing child sexual abuse investigations using biometrics and social network analysis. Trends and Issues in Crime and Criminal Justice, 668, 1-16. https://doi.org/10.52922/ti78948 [PDF]
  • Westlake, B., Brewer, R., Swearingen, T., Ross, A., Patterson, S., Michalski, D., Hole, M., Logos, K., Frank, R., Bright, D., & Afana, E. (2022). Developing automatic methods to detect and match face and voice biometrics in child sexual abuse videos. Trends and Issues in Crime and Criminal Justice, 648, 1-15. https://doi.org/10.52922/ti78566. [PDF]
  • Brewer, R., Westlake, B., Hart, T., & Arauza, O. (2021). The ethics of web crawling and web scraping in criminological research: Navigating issues of consent, privacy and other potential harms associated with automated data collection. In A. Lavorgna & T. Holt (Eds.), Researching cybercrimes (pp.435-456). Palgrave Macmillian: Cham. https://doi.org/10.1007/978-3-030-74837-1_22 [PDF]
  • Westlake, B. G. (2020). The past, present, and future of online child sexual exploitation: Summarizing the evolution of production, distribution, and detection. In T. Holt & A. Bossler (Eds.), The Palgrave handbook of international cybercrime and cyberdeviance (pp.1225 1253). Palgrave Macmillian, Cham. doi: https://doi.org/10.1007/978-3-319-78440-3_52. [PDF]
  • Westlake, B. G. (2018). Delineating victims from perpetrators: Prosecuting self-produced child pornography in Youth Criminal Justice Systems. International Journal of Cyber Criminology, 12(1), 255-268. doi: https://doi.org/10.5281/zenodo.1467907. [PDF]
  • Westlake, B. G., Bouchard, M., & Frank, R. (2017). Assessing the validity of automated webcrawlers as data collection tools to investigate online child sexual exploitation. Sexual Abuse: A Journal of Research and Treatment, 29(7), 685-708. doi: https://doi.org/10.1177/1079063215616818. [PDF]
  • Westlake, B. G., & Frank, R. (2016). Seeing the forest through the trees: Identifying key players in online child sexual exploitation distribution networks. In T. Holt (Ed.), Cybercrime through an interdisciplinary lens (pp.189-209). New York: Routledge. doi: https://doi.org/10.4324/9781315618456. [PDF]
  • Westlake, B. G., Bouchard, M., & Frank, R. (2012, September). Comparing methods for detecting child exploitation content online. In 2012 European Intelligence and Security Informatics Conference. Odense, Denmark, 22-24 August (pp. 156-163). IEEE. https://doi.org/10.1109/EISIC.2012.25. [PDF]
  • Joffres, K., Bouchard, M., Frank, R., & Westlake, B. G. (2011, October). Strategies to disrupt online child pornography networks. In 2011 European Intelligence and Security Informatics Conference. Athens, Greece, 12-14 September (pp. 163-170). IEEE. https://doi.org/10.1109/EISIC.2011.32. [PDF]
  • Frank, R., Westlake, B. G., & Bouchard, M. (2010, August). The structure and content of online child exploitation. In ISI-KDD ’10 ACM SIGKDD Workshop on Intelligence and Security Informatics, Washington, USA, 25-28 July (Article 3). ACM. https://doi.org/10.1145/1938606.1938609. [PDF]

Grants

  • 2022-2023: Advancing child sexual abuse victim and perpetrator identification using biometrics (P21/519, Australian Institute of Criminology, Commonwealth of Australia)
  • 2021-2022: Developing automated methods for detecting child sexual abuse videos online (C25-00210, Cyber Security Cooperative Research Centre, Commonwealth of Australia)
  • 2020-2020: Validating an automated biometrics tool for detecting child sexual exploitation in videos (Central RSCA, San Jose State University)
  • 2019-2021: Developing automated audio and facial recognition biometric tools for detecting child exploitation material (P18-320, Australia Institute of Criminology, Commonwealth of Australia)
  • 2012-2015: Monitoring the evolution of online child exploitation networks (Insight Grant, Social Science and Humanities Research Council)
  • 2012-2015: The criminal career evolution of child exploitation websites: Identification, survival, and community patterns (767-2012-2438, Joseph-Armand Bombardier Canada Doctoral Graduate Scholarship, Social Science and Humanities Research Council)