Since its inception, social media has evolved to become one of the greatest tools in the modern marketer's arsenal. But, while companies use this emerging medium to disseminate brand messaging and conduct customer service, individual consumers use social media as their personal mouthpiece, sounding off about whatever's on their mind. Yet, while predictive analytics have become the ideal companion tool for marketers looking to gain granular insights into their social consumer base, such technologies have begun to bleed into other areas of our society, impacting the way law enforcement officials conduct business.For the Detroit Crime Commission (DCC), social now acts as its peephole into the criminal mind. The DCC recognized that many criminals were posting about their crimes across various social media platforms, announcing potential plans, flaunting drugs and weapons on Facebook, Twitter, and Instagram, and organizing their next move. However, by making such information transparent to the public, the DCC decided to take advantage of this open data by partnering with Semantria to introduce text analytics that would allow the team to track criminal elements, activities, and consequences.
Together, the DCC and Semantria created a custom gang-slang dictionary, allowing the DCC to monitor and mine real-time data from Twitter and Facebook using keywords such as kill, shoot, shank, and murk. "We worked together to build a custom dictionary specific to Detroit slang," says Oleg Rogynskyy, CEO of Semantria. "This way, our engine could pick up on the local slang and other obscure words. Now, when the DCC runs data from social media through Semantria, the output consists of only tweets and comments that are related to criminal activity, while everything else would be filtered out."
Since its implementation, the DCC has seen the amount of criminal activity on social media in Detroit decline greatly, with a noted decrease in overall crime throughout the city. Rogynskyy notes that, while social media allows people to connect quickly and easily, making said platforms ideal for facilitating crime, law enforcement officials can take advantage of the same platforms by using text analytics to predict what may happen next before anyone gets hurt. "They can see things such as who is going to commit the crime and where the crime is going to happen, and this allows them to prevent crime before it happens," Rogynskyy adds. Such tools put the power in the hands of those designated to protect the community, enabling them to do their job better than ever before.