An Algorithm to Pinpoint Human Traffickers

May 4, 2021

We love applications of machine learning that actually benefit society. Here is one that may soon be “Taking Down Human Traffickers Through Online Ads,” reports the Eurasia Review. The algorithm began as a way to spot anomalies (like typos) in data but has evolved into something more. Now dubbed InfoShield, it was tweaked by researchers at Carnegie Mellon University and McGill University. The team presented a paper on its findings at the most recent IEEE International Conference on Data Engineering. We learn:

“The algorithm scans and clusters similarities in text and could help law enforcement direct their investigations and better identify human traffickers and their victims, said Christos Faloutsos, the Fredkin Professor in Artificial Intelligence at CMU’s School of Computer Science, who led the team. ‘Our algorithm can put the millions of advertisements together and highlight the common parts,’ Faloutsos said. ‘If they have a lot of things in common, it’s not guaranteed, but it’s highly likely that it is something suspicious.’”

According to the International Labor Organization, ads for four or more escorts penned by the same writer indicate the sort of organized activity associated with human trafficking. The similarities detected by InfoShield can pinpoint such common authorship. The organization also states that 55% of the estimated 24.9 million people trapped in forced labor are women and girls trafficked in the commercial sex industry. Online ads are the main way their captors attract customers. The write-up continues:

“To test InfoShield, the team ran it on a set of escort listings in which experts had already identified trafficking ads. The team found that InfoShield outperformed other algorithms at identifying the trafficking ads, flagging them with 85% precision.”

The researchers ran into a snag when it came to having peers verify their results. Due to the sensitive nature of their subject, they could neither share their data nor publish examples of the similarities InfoShield identified. Happily, they found a substitute data sample—tweets posted by Twitter bots presented a similar pattern of organized activity. We’re told:

“Among tweets, InfoShield outperformed other state-of-the-art algorithms at detecting bots. Vajiac said this finding was a surprise, given that other algorithms take into account Twitter-specific metrics such as the number of followers, retweets and likes, and InfoShield did not. The algorithm instead relied solely on the text of the tweets to determine bot or not.”

That does sound promising. We hope authorities can use InfoShield to find and prosecute many, many human traffickers and free their victims.

Cynthia Murrell, May 4, 2021


Got something to say?

  • Archives

  • Recent Posts

  • Meta