Analysing child sexual abuse activities in the dark web based on an efficient CSAM detection algorithm
Ngo, Vuong M.ORCID: 0000-0002-8793-0504, Thorpe, ChristineORCID: 0000-0002-2359-883X and Mckeever, SusanORCID: 0000-0003-1766-2441
(2023)
Analysing child sexual abuse activities in the dark web based on an efficient CSAM detection algorithm.
In: 2nd Annual Trust and Safety Research Conference, 28-29 Sept 2023, Stanford,CA, USA.
(In Press)
Child sexual abuse material (CSAM) activities are prevalent on the Dark Web to evade detection, posing a global challenge for law enforcement. Our objective is to analyze CSAM discussions in this concealed space using a Support Vector Machine model, achieving an accuracy of 87.6%. Across eight forums, approximately 28.4% of posts contained CSAM, with victim ages most commonly reported as 12, 14, 13, and 11 years old for YouTube, Skype, Instagram, and Facebook, respectively. Additionally, in forums discussing boys, the most frequently mentioned nationalities in CSAM posts were English, German, and American, accounting for 12%, 7.8%, and 6% of all nationalities, respectively.
Metadata
Item Type:
Conference or Workshop Item (Paper)
Event Type:
Conference
Refereed:
Yes
Uncontrolled Keywords:
CSAM analysis; supervised learning; Dark Web post; social media and
mental health
N-Light project which is funded by the Safe Online Initiative of End Violence and the Tech Coalition through the Tech Coalition Safe Online Research Fund (Grant number: 21-EVAC-0008-Technological University Dublin).
ID Code:
29282
Deposited On:
15 Dec 2023 12:02 by
Vuong M Ngo
. Last Modified 15 Dec 2023 12:02