Exposure
AI-Powered Dashboard Targeting Online Human Trafficking.
Real-Time Monitoring
Continuously tracked suspicious online human trafficking activity in real time.
The Opportunity
‘Of the 359 sex trafficking cases Toronto police have investigated since 2013, every single girl was advertised on Backpage.com… ‘ a now defunct online sex for sale site. At the time the massive growth in online adult listings, coupled with static social service and enforcement resources, made reaching those who critically needed support a near impossible task.
The Approach
In collaboration with the City of Edmonton’s Community Safety Supervisor, I developed the tech for Project Exposure, an NLP-powered dashboard designed to detect potential underage human trafficking activity online. Leveraging advanced analytics, keyword monitoring, and custom algorithms, the dashboard aggregated data to identify suspicious patterns and provided actionable insights for immediate intervention.
The Impact
Project Exposure empowered bylaw officers assigned to monitor both online and offline trafficking activity. They went from reviewing only a fraction of Adult Listings for infractions and at-risk individuals, to covering 100% of relevant online listings, ensuring comprehensive surveillance and enabling officers to focus on critical interventions.
Features
Vision AI Age Determination
Identified approximate ages using advanced AI algorithms.
Comprehensive Dashboard
Aggregated and visualized key insights for actionable intelligence.
Leveraged TOR Networks
Scraped data through TOR networks to avoid target site blocking tools.
Technology
- Front-end: HTML, CSS, JavaScript (jQuery, w2ui, vis.js Timeline, Semantic UI)
- Back-End: VM, Ubuntu, PHP, MySQL, Tor.exe
- Analytics with Python (BeautifulSoup, nltk)
Disclaimer
Some sensitive data elements in the screenshots have been obfuscated to ensure privacy.
