This blog post is an excerpt from GovLoop’s recent guide, How You Can Use Data Analytics to Change Government. Download the full guide here.
The data analytics landscape is rapidly evolving thanks to more powerful and affordable tools for data gathering and processing. Memory, CPUs, and fast storage are getting cheaper as the manufacturing processes are improved. Each generation is becoming more familiar with the ecosystem of analytics and data storage technology, said Billy Hill, Customer Solution Architect for Big Data Analytics at ViON. This combination of affordability and usability bring analytics tools within reach for government agencies, businesses and even consumers.
“Although the technology has been around for years, it’s now accessible to a wider audience,” Hill said.
People are using more CPUs and memory in concert, and these technologies are starting to be incorporated into the workflow of researchers and mathematicians, who ultimately develop the algorithms used to find patterns in the data. Advances in hardware are helping to drive changes in the types of algorithms that are developed. At the same time, the algorithms are also dictating how the hardware is used. “It’s a synergistic development,” Hill said. As the technology advances and provides better insights, the demand for data analytics and the outcomes they produce continues to increase. “Leaders nowadays are expecting to get more value out of their data,” Hill said.
That’s why agencies are using different types of analytics, such as predictive and prescriptive to better understand what may happen in the future and how they should respond. Depending on the type of analysis they want to conduct — real-time or historical — there are different data gathering and analytics tools to support those efforts. Let’s take a closer look at each of these examples to better understand what they entail and how they benefit government agencies.
Predictive and prescriptive analytics. It’s hard to say with certainty what will happen a month from now or even a year from now. But agencies can use the data they have to better predict what may happen in the future. That’s the real value of predictive analytics. It helps agencies plan accordingly in a variety of areas, whether it’s budgeting, future customer demands for services or internal human resource planning. But knowing what may happen does an agency no good if it’s unprepared to respond. That’s where prescriptive analytics comes into play.
“If agencies are able to predict the probability of an event or action occurring, that information can be used to drive their decisions,” Hill said. “Prescriptive analytics helps decision- makers figure out what to do with the results from their models. To make decisions you have to be informed, and the better your information, the better your decisions.”
Streaming vs. batch analytics. Streaming and batch refer to the types of analytics technologies that are used to drive specific outcomes. For example, streaming technologies are usually used to power real-time data analysis that require immediate data processing and insights. Streaming technologies involve the use of sensors to rapidly collect the data that agencies must then analyze.
Batch analytics, on the other hand, is used in scenarios that don’t require immediate decision- making. In this instance, agencies have more time to tweak their models for more accurate insights. “There is a greater focus on quality if you’re doing batch, where as with real-time analysis there are online algorithms making decisions as they’re running,” Hill said.
Real-time vs. forensic analysis. Now that we’ve discussed some of the types of analytics agencies can use, let’s explore the outcomes they produce. To reiterate: Streaming analytics supports real-time analysis, while batch analytics supports forensic analysis.
Both real-time and forensic analyses provide great value to agencies, depending on their specific requirements, Hill said. Real-time analysis is helpful when decisions are based on quickly evolving data, such as current traffic on the road or changing weather conditions. But batch analytics would be the more appropriate option for driving forensic analyses looking at historical data, such as traffic patterns over time or weather conditions over the past week or several months.
In an emergency response situation, first responders would rely on real-time analysis to determine how far an ambulance is from a victim, and how far the victim is from the nearest hospital. A forensic analysis of that same data may be used to better understand which parts of the city receive the most 9-1-1 calls and where ambulances should be stationed to better respond to those calls. “We’re always interested in things that are just out of reach, and some people think this technology is out of reach — but it’s not,” Hill said. Thanks to data analytics, agencies have all of these capabilities and more at their fingertips.