Organizations of all kinds are applying analytics to a wide range of activities. Analytics are used to measure performance and productivity, gather insight into behavior, and search for common threads or links between seemingly unrelated events. Basically, analytics are a way to identify patterns in data, and when performed with powerful computer programming and the right algorithms, these patterns can provide real insight into what is happening over time and in specific situations--even in real time.
One of the most common applications of analytics is to monitor the many video channels surveilling a given area. But today, the scope of analytics is much wider and is often referred to as "content analytics". Content includes any form of communication that can be captured or recorded. Today, this includes video, speech, and text.
As technology has matured, law enforcement has begun utilizing analytics in a variety of ways.
Video analytics
In security and public safety, video analytics are becoming an ever-more reliable source of information with increased accuracy and reduced false alarms when used in real time. They are also increasingly being used for post-event and near real-time forensic analysis to find subjects of interest quickly. The accuracy of video analytics has substantially increased, in correlation with our understanding of how they work in situ and today’s processing power. For example, environmental influences, such as lighting or weather, that once gave off false alarms can now be filtered out. Increased accuracy combined with the ubiquitous nature of video surveillance has enabled the development of new analytics applications.
Most cities today are covered with video surveillance cameras; some have more than a million video channels in operation, recording everything happening in their view. But just because something or someone has been recorded on video doesn’t mean finding that particular incident or person in the mounds of video recordings is an easy task--that is, until now.
New, patent-pending video analytics technologies make locating and tracking a specific person from surveillance video faster, minimizing search time from hours to minutes.
Instead of manually scanning hours of video, a task that could take hours or days to complete, with no guarantee of accurate results, law enforcement can tell this "smart software" who or what to look for, using either an uploaded picture or a user-generated avatar. The software will scan the video and identify relevant matches within seconds. Law enforcement can continue to refine these results until a match or several close matches are identified. All of this takes place over the span of several minutes. Moreover, once the results are identified, officers can then track the suspect’s location, based on the location of the cameras he was spotted by. The key to providing super-fast search results is indexing the video while it is being recorded.
Beyond reducing the time and resources used to find a suspect, this new technology allows law enforcement to substantially mitigate the risk stemming from the suspect’s actions, since he can be apprehended quickly. Criminal suspects are not the only individuals that can be tracked with this technology; there are other applications as well. For example, law enforcement can use it to search for and find missing children within a surveillance network. When time is of the essence, analytics applications can have a profoundly positive effect on the outcome of an incident.
Speech analytics
Speech analytics has also improved over the last few years and has become a sought after technology for analyzing emergency center communications. Speech analytics, or speech indexing, lets you search for spoken words within recorded audio. With it, law enforcement and more specifically Public Safety Answering Points (PSAPs) can quickly find and link calls wherein the data provided by the caller is not sufficient to identify them. This can be a very useful tool to support investigations, ensure regulatory compliance, and perform quality assurance.
Now let’s say a PSAP wants to aggregate calls over a specific time period. Using keywords and topics, speech analytics can identify all of the calls that are related to that particular subject, which can then uncover patterns relating to a range of parameters--location, incident type, time of day, week or month, and more.
Speech analytics can also help ensure regulatory compliance and quality assurance. Users can search for specific calls that might have required special handling and then check to make sure that regulations and policies were followed. This can often mitigate liability, and provide insight into knowledge gaps that must be addressed, and uncover best practices to be implemented.
Text analytics
Text analytics have become increasingly relevant since PSAPs began adopting text-to-9-1-1. With text analytics users can search for keywords and phrases to bring up all communications related to a particular subject matter, much in the same way that speech analytics operates. Even text response time can be analyzed and provide telling information.
During an investigation, law enforcement can analyze text to reveal patterns that could link communications to their sender. It can also identify potential risks through sentiment analysis.
Correlating information from analytics
Video analytics, License Plate Readers (LPR), and Crowd and Intrusion Detection applications are nothing new, but when used with technology such as Physical Security Information Management (PSIM) platforms, are increasing their value by correlating the insights these applications extract. For example, say an LPR registered a speeding car in the same vicinity in which an intrusion detection alert was raised just moments prior. Certain PSIM technology has the ability to correlate these two seemingly separate incidents, giving law enforcement insights they otherwise wouldn’t be able to attain so quickly or easily.
On their own, speech and text analytics also have the ability to link incidents. Let’s say a 9-1-1 call center receives several communications about an incident, both via telephone call and text message. Each caller or communication provides relevant and distinct information about the incident. Analytics are able to immediately flag these multiple communications as being related to a single incident. This gives law enforcement a more complete picture of the event, allowing for a more effective response.
The Value of analytics
With so much data flooding control rooms these days, law enforcement needs to be able to distill the massive amounts of unstructured content into relevant information. Video, speech, and text analytics help do that by linking separate incidents or calls and generating insights about incidents that would take much longer to uncover, if ever.
Analytics can be applied not only to improve response times and investigations, but also to ensure that performance and regulatory standards are upheld. The amount of information available is only going to increase, and analytics help make that information as valuable as it can be.
