|
Abstract
|
Technologically enabled crime has proliferated in recent years. One such crime type is the live streaming of child sexual abuse (CSA). This study employs a machine learning approach to better understand the characteristics of Australians who engaged with known facilitators of CSA live streaming in the Philippines.
This model demonstrated notable success in identifying the individuals who would engage in a high number of transactions with known facilitators.
Individuals engaged in high-volume live streaming typically spent small amounts (under $55) at intervals of less than 20 days. Where prolific offenders had a criminal record, it was unlikely to consist of high-harm crime types, such as violent or sexual offences.
This study considered whether machine learning analytics could offer insight into the transaction and offending behaviours of prolific live streamers of CSA. This is an emerging body of work in which the characteristics of offenders are largely unknown. The frequency and monetary value of transactions among these individuals are particularly important and have implications for identifying these crimes among financial transactions data. Offenders did not appear to have engaged in violent offending; rather, a history of low-harm offending was common, although the under-reporting of sexual offences among children and adults is an important consideration.
FINDINGS here may contribute to a better understanding of and ability to identify offenders who pay to watch the abuse of children via live stream, but further analyses of this type employing a control group could substantially contribute to this area of work.
Keywords: Human trafficking;
Language: en |