Human traffickers could be identified through their Bitcoin transactions
Image credit: Dreamstime
US researchers have collaborated to develop two machine learning algorithms which could help law enforcement identify human trafficking rings advertising exploitative sexual services online through analysis of linguistic style and linking adverts to Bitcoin transactions.
Human trafficking is a devastating international problem which affects every country in the world. According to the International Labour Organisation, an estimated 21 million people are living as slaves, of which 4.5 million are sex slaves.
Criminal groups profiting from sexual exploitation can make the most of the lack of regulation online to place semi-anonymous adverts for sexual services. This veil of anonymity causes a headache for law enforcement services attempting to identify these groups, which often hide behind many different contact details in order to avoid being tracked down.
“There are hundreds of thousands of these ads placed every year, and any technique that can surface commonalities between ads and potentially shed light on the owners is a big boost for those working to curb exploitation,” said Professor Damon McCoy, a New York University assistant professor of computer science and engineering.
Researchers from the University of California-Berkeley, New York University and the University of California-San Diego collaborated on a project to develop two free machine learning tools to help police and non-profit groups identify victims of human trafficking hidden behind adult adverts.
The first algorithm is based on stylometry: the in-depth analysis of linguistic style to identify authors across multiple texts, which are often anonymous or fraudulent. When applied to online sex adverts, this tool can identify a single author behind multiple adverts on Backpage (a classifieds advertising website popular for sexual services): a warning sign of a human trafficking ring. This process can be automated, allowing for the quick identification of cases which could invite further investigation.
“Imagine going through page after page of explicit advertisements, some for underage victims,” said Rebecca Portnoff, a PhD student in computer science who developed the algorithm. “You’re looking through all this material to find the set that are advertising trafficked and underage victims. Even given a team of humans dedicated to the task, there’s simply too much data – often quite traumatising – to go through.”
The second approach developed by Portnoff and her colleagues is particularly effective once the first algorithm has identified a single author behind multiple adverts. This algorithm is based on the fact that Bitcoin is the primary payment method for online sex ads. Records of all users’ transactions completed with the virtual currency are stored in the Bitcoin “mempool” (memory pool) and blockchain, and are publicly available.
Backpage posts adverts as soon as payment is received, allowing the researchers to compare the timestamp of payment submission in the transaction records to the timestamp of the ads appearing on the website. By tracing a paid advert to a unique Bitcoin user’s account, you can identify individuals or groups involved in human trafficking.
The researchers tested their automated identification methods on 10,000 adverts on Backpage, as well as several that they wrote and submitted themselves for comparison. The researchers found that their methods were 89 per cent successful at grouping adverts by author – a considerable improvement on current methods – and also effective at linking their own ads with their accompanying blockchain timestamps. They were unable to find, however, whether the users identified could be tied to human trafficking, a matter for law enforcement to pursue.
“The technology we’ve built finds connections between ads,” said Portnoff. “Is the pimp behind that post for Backpage also behind this post in Craigslist? Is he the same man who keeps receiving Bitcoin for trafficked girls?”
“Questions like these are answerable only through more sophisticated technological tools – exactly what we’ve built in this work – that link ads together using payment mechanisms and the language in the ads themselves.”
In recent months, Bitcoin has soared in value to nearly £3,500. Canadian financial regulators have recently endorsed the currency: the first official endorsement of money created independently of the government of central banks, and Egypt is due to open its first Bitcoin exchange in August 2017.