X
Business

How the police use social media to catch crooks and make friends

Businesses are discovering they need to be where their customers are, but rather than gaining or losing a sale, the police have realised that this is their opportunity to stop crimes and build community relationships.
Written by Michael Lee, Contributor

Law enforcement agencies are beginning to use social media to catch crooks, but also to change their methods of policing.

Former Singapore Police officer turned Oracle global lead for justice and public safety Hong-Eng Koh explained to ZDNet how police are now having to adapt their policing methods.

He said that there was a large shift in behaviour in society that he predicted would result in a decrease in the number of emergency calls being made.

"In the old days when you saw a bank robbery, you would call 000. But today ... we are predicting the number of emergency calls to go down because today the Generation X and Y, when they see a bomb explosion, or a big fire, what's the first thing they do? Tweet it. They take a photograph and tweet it. They won't call, they want to share among their friends."

It has meant that law enforcement organisations have had to up their game when sourcing information, going to Twitter and other social networks to find information before it gets reported.

Using social networks as a source of information is possible as Koh estimates that 40 percent of social media users do not make their contributions private. The difficulty of using such data, however, is the unstructured nature of it.

Making sense of all the data is possible by using several techniques. Koh said that law enforcement has used pattern and event recognition to first identify topics or keywords that may be of significant interest — in other words, a law enforcement-tailored way of detecting trending topics.

These can be formed into rules and set to trigger an alert if movement on a particular topic exceeds a certain pattern.

Knowing that the public is talking about a particular issue is not enough, however, so according to Koh, an additional tool law enforcement uses is sentiment analysis.

"Are they happy or unhappy? That's important from a police perspective because if a big group of people are unhappy, I'd better be careful that something is going to happen," he said.

"We're not going to be intelligence-led by something that has already happened. We are more focused on preventing."

Further metadata can be used to determine and filter out information to make it more relevant to policing needs. Koh said that as the Ya'an earthquake took place earlier this year, it was possible to identify individuals that needed urgent assistance, and primary sources of information using the above techniques, coupled with the geolocation data attached to many tweets.

Poring over social networks for intelligence is not the only use that the police have for the practice, however. Koh said that more and more, law enforcement organisations were beginning to go back to their roots where community police officers were among the people and built relationships. He said that policing had gone through a rough patch where they were seen as custodians as power, rather than those among the people, and that social media was a way to return to what is called community policing.

"Look at Japan. They have the Koban system — they build many police posts throughout Japan, the police officers stand in front of the post, and they interact with the community. Instead of calling 911 or 000, the people come to you."

He said that younger people are online, however, and the police need to be where the people are.

Koh pointed to the example of the Boston Marathon Bombing earlier this year, in which the public was asked to come forth with any information.

"Boston Police actually brought themselves down on the ground level, as a friend [and said], 'Guys, I need help. I know you guys have been taking photographs, video out there can you send it to us?'"

"If Boston Police Department did not have that good relationship built up through social networking before that, they probably would not have received such a great outcome."

Editorial standards