memex iceberg

Fighting Crime in Cyberspace: how Spicule became part of the data police

Picture a group of vicious criminals finally being apprehended and brought to justice. You are probably visualising an FBI SWOT team, flashing lights and a dramatic stand-off. You may even be hearing some frenetic, bass-heavy soundtrack too, thanks to a steady diet of Netflix crime series. What you won’t be imagining is a team of developers, huddled over their keyboards, creating a platform designed to bring the bad guys down. Yet much of modern crime-fighting is data-driven, and that’s how Spicule got involved.

Two years ago, our founder Tom Barber started working as part of a team tasked with using data solutions to track down criminal activity. Their challenge was to create a platform that could be used to scan the web (including the dark web) to accurately identify illegal dealings and help the FBI to build a picture of where these operations were taking place. As many of the Bureau’s agents aren’t tech experts, the platform needed to be user-friendly and easy to navigate.

Avengers Assemble

Tom was recruited by NASA JPL, a federally funded research and development centre based in California, as part of their team of data experts tackling this vital work. JPL was joined on the project by data professionals from NIST (National Institute of Standards and Technology) and representatives from several leading US universities. The group was managed by DARPA (The Defense Advanced Research Projects Agency) as they began looking into a number of different criminal domains – including human trafficking, weapons smuggling, child exploitation, drug smuggling and financial fraud. By bringing some top minds in data analytics together to build the platform, DARPA was determined to make a significant impact on hunting down these criminals in cyberspace.

The Justice League

The project focused on scanning the web for adverts and trying to decipher if those posted were legitimate or a front for offenders. In many different places, services such as escorts are legal, but the adverts posted on sites like Craig’s List may also be used to offer up human trafficking or child exploitation. Differentiating between legitimate advertising and illegal posts is tricky as the differences can be incredibly subtle. But by using machine learning, the team was able to begin teaching the platform to spot anomalies.

The machine learning job involved feeding lots of existing data into the platform until it began to identify the differences. From that point on, the team could add in new data and then cluster ads together that seemed to comply with learning on criminal activities. Interestingly, thanks to smartphone technology and photo tagging, many images used in these clusters still had GPS coordinates on them, as well as other metadata. From this information, the FBI could narrow down and begin to focus in on groups of interest, helping them to create a targeted response.

Guardians of the Galaxy

There were no capes or fancy badges involved but being part of this project was one of the most fulfilling jobs that Tom has ever done. Today, elements of the platform have been taken away by different law enforcement agencies – including the Metropolitan Police – to help them protect the victims of these crimes.

“Work like this changes your outlook,” explains Tom. “It means a lot to know that we’re part of a project that is helping to improve people’s lives across the world. Working with NASA JPL was already a dream come true for me, but when that also involves doing something that will make a difference to others, it’s a great feeling.”

If you’d like to learn more about Spicule’s data analytics work and how it could be applied in your business, please get in touch on 01603 327762 or email info@spicule.co.uk.