Safe Neighborhoods with Big Data
When I was coming back from an international travel recently and was browsing through the inflight entertainment shows available, "Person of Interest" series on CBS caught my eye. A computer genius develops a machine for the government which is used to detect information leading to acts of terrorism before they can be executed. "The machine" separates the information gathered into two categories: relevant and irrelevant. The scientist, however, discovered that the irrelevant information also often led to the discovery of other acts of violent crimes. With the state-of-the-art surveillance technology, the program uses pattern recognition to identify people who are about to be involved in violent crimes.
The high-tech tools that are used in the series are more factual than fiction. That sort of facial recognition to identify a person in a sea of faces is not sci-fi. In recent times similar technology was used to identify people who took part in the London Riots.
"The Machine" in the series is nothing but the Big Data, which the local law agencies collect and store. Today, many law enforcement agencies are participating in data sharing initiatives in which local data captured is being provideded to various regional and federal data warehouses. This aggregated data is used for search and analytics functions enabling data to be delivered quickly to law enforcement officers with the need to know.
We in the public are being looked at; we are being listened to, tracked by FastTrak we use on the bridge; We can be tracked globally by satellites via mobile phone companies; We leave ‘breadcrumbs’ in our Twitter, Facebook and other social network communications. A data scientist can identify all these patterns using big data analaytics and can help us create a safe and secure neighborhood.
To maximize the use of this Big Data, a combination of data interface standards, flexible BI software and proper human capital (e.g. a Data Scientist) is required. Data interface standards will ensure the proper and consistent retrieval of big data. BI tools need to support industry standards such as Hadoop, which make it easier to work with Big Data.
Such capability needs to be implemented across all the local law enforcement agencies (at least at a regional level) and governments should support Data Scientists who can analyze the data accessible through social networking sites, video surveillance, blogs and protect the public from the possible crimes happening in the city and prevent crime. Especially around the universities and schools where there could be a person who suddenly picks up a gun and randomly shoots at innocent people, as we recently saw at Chardon, Ohio school shootings. Perfect example is the Predictive Policing method which Santa Cruz Police Department used as an experiment in July last year helped officers pre-empt several crimes and has led to five arrests. In the current economical conditions, where Police Departments are having big budget cuts and downsized staff, these effective alternate methods will enable the departments to focus on crime prone areas, predict and stop crimes.
I think having the Big Data and funded Data Scientist resources would help the community and the nation enhance safety.
References
1) Person of Interest Video Clip at CBS
2) London Riots Facial Recognition Technology
3) Santa Cruz and Los Angeles Predictive Policing 6 month Trail
4) Chardon, Ohio School shooting
Further Reading
1) Saama SixthSense - Download a point of View Presentation
2) Saama’s Data Scientists -
3) Harvester - Saama's Solution to simplify, automate and accelerate the entire process of data integration.