IT professionals hear a lot about Big Data, but perhaps an even more useful concept is smart data. State and local governments that use data strategically can greatly improve their decision-making and boost efficiency. Machine learning is an emerging technology that can help agencies intelligently use the data they produce and collect.
SAS defines machine learning as "a method of data analysis that automates analytical model building." This approach to artificial intelligence enables computers to learn through experience. State and local governments haven't waded deeply into machine learning yet, but many are starting to dip their toes into the waters of the technology.
Many local governments — possibly more than you think — are finding innovative ways to implement machine learning for better decision-making, improved security and streamlined automation. In fact, in its 2017 Digital Cities Survey, the Center for Digital Government found that 64 percent of cities are using or plan to use predictive analytics and machine learning. And in a recent report, the National Association of State CIOs found that 29 percent of state CIOs expect artificial intelligence/machine learning to be the most impactful emerging technology in the next three to five years.
Machine Learning Bolsters Public Safety Productivity
The potential use cases are nearly limitless for state and local agencies looking to implement machine learning, but perhaps the most common use is to improve decision-making.
The Sacramento County Sheriff's Department in California, for example, uses artificial intelligence tools to analyze crime statistics and other information to identify trends and improve police operations, according to a Splunk customer case study. This implementation supports proactive policing and saves the department hundreds of hours per year in ad-hoc report generation.
"Before Splunk … There was no integration among our key systems. We had to either write an SQL script or use the SQL Server Reporting Services report generator to produce each report," John Britto, senior IT analyst and application team lead for the County of Sacramento, notes in the case study. "Our command group is now able to more clearly see trends in our crime statistics and take proactive action to address areas of concern and provide the best possible service to the public."
Video Analytics Enhance Law Enforcement
Video analytics is another use case for which machine learning can deliver significant value, especially in law enforcement.
Surveillance cameras, traffic monitors and police body cameras produce massive amounts of data each day — far more than could be analyzed by human users. However, machine learning tools can take on this workload and make sense of the vast piles of data produced.
"Video analytics can greatly assist law enforcement and public safety by revolutionizing how video and multimedia data are searched, tagged, used and managed," Stephen Russo, IBM's director of cognitive city solutions, states in a blog post.
Machine learning tools can interpret and index the video captured by a camera, such as compiling appearances of persons who match a particular description or instances in which individuals engage in a specific activity. They can also alert users to behaviors, such as when someone enters a restricted area or leaves a bag unattended. This analysis can provide law enforcement officers with valuable insights, such as situational awareness and real-time assessments.
"Today's mainstream dialogue around body-worn cameras is focused only on eyewitness accounting and the costs associated with storing and managing requirements for video. Realizing that the value of the video captured in this manner is not just in capturing it, but also in finding and using what is in the footage is equally important," Russo adds.
Data Security Gets a Boost with Machine Learning
State and local agencies also can use machine learning to improve their data security, Forbes reports. It's a massive challenge for anti-virus software vendors to keep up defenses against the large number of variations of malware that appear in the wild. But these variations are only marginally different from the original malware, and machine learning solutions can detect them.
Other security tools use machine learning algorithms to identify patterns in data traffic and report anomalies that could indicate a security breach.
As government agencies grow more familiar with machine learning technologies, they can expect to expand their use of them.