What Is Process Mining, and How Can It Help Governments Transform?

What Is Process Mining?

As ABBYY defines it, “process mining is an approach to process improvement that’s driven by advanced data analytics. Process mining software uses process event log data to discover and map process operations and their possible variations. These analytics help decision-makers uncover the root causes of any inefficiencies in their business processes to reveal opportunities for improvement.”

To optimize processes, state and local agencies can rely upon data to inform their decisions in allocating resources. Process mining might begin with an audit of data logged by enterprise resource planning and customer relationship management tools, for example.

If an agency or department is beginning (or planning) to automate elements of work, process mining can be a leading tactic to make work best at whatever scale is needed. 

DISCOVER: How can artificial intelligence and other technologies help smart cities share data?

What Is Conformance Checking in Process Mining?

One element of process mining that can get lost in discussion is conformance checking, which is a technique used to compare event logs or an actual process with the existing reference model (or target model) for that process. 

Often, this is referred to as finding a “happy path” — when an input produces the expected output. 

It is important to note that every happy path is different, Raiker says. To think of it in a nontechnical sense, remember that one size doesn’t always fit all. Raiker cites, as an example, an audit of a customer’s process that required 16 full-time employees. 

“They had only about 8 to 10 percent of all of the workflows audited. They simply couldn’t complete the full audit,” he says. “So, they going to use a sample size to try to understand the results, but that would be noncompliant. And also — perhaps in many of those cases that go unseen — it would be nonconforming. In using ABBYY’s process intelligence solution, they were able to load up all of the workflow.” 

That process didn’t necessarily follow the traditional path, but it did lead to the customer’s defined happy path. 

“They had no other way,” Raiker says. “There’s not another technology along business intelligence or Excel spreadsheets that can understand this, but process intelligence can.” 

The customer was able to shrink its audit team of 16 full-time employees by nearly half, sending the rest to another audit. 

“Here’s another process area that we want to get more information about,” Raiker says. “We want to be more conforming. We want to be closer to compliance and the regulations, but we don’t have a simple solution. We’re going to move our team.”

How Can Process Mining Help State and Local Government?

According to Raiker, a lot of the issues found when working with governments are discovered in the procurement process. Yet, a lot of that work isn’t what citizens encounter or what causes them frustration. 

“It could be the streets department, the sanitation department, the water department,” he says. “What we see with digitalization is that it allows deeper communication and collaboration across these traditionally siloed departments. 

“So, what process mining can do is help highlight where systems are working, where things are going OK, building out structures and process documents that can help you automate some of the routing and procedural things.”

LEARN: How to solve for technical debt in state and local government agencies.

He notes that process mining also gives state and local governments the ability to zoom in on problem vehicles, departments and workers.

“We’re not looking to highlight anyone and say, ‘This person did it badly,’” Raiker says. “But why is it that there’s a very specific firetruck or ambulance that takes far longer to respond to calls than another?” 

Process mining offers the potential to not only learn the answer but also to solve for it. 

Process Mining Is Dependent on Interagency Information Sharing

Information is at the core of what governments do. Whether that information is something that goes directly back to a citizen or is used to improve a service, governments need information and can thrive with it.

“Information sharing is a positive in these scenarios, and that’s the disruptor,” Raiker notes. “Process mining enables information. Partially, it’s information that was never able to be seen before, about how one department interacts with another.” 

READ MORE: Virginia's new chief data officer has big plans.

Raiker describes a hypothetical scenario in which an employee, who works in a state revenue department, files a form electronically. Once the form leaves that employee’s system, it goes to three additional destinations. From there, it may go somewhere else. Another employee in the state accounting office may eventually receive it. 

An examination of the log data from that transaction could reveal information about the governance of an agency or a program. Even knowing this much is part of a potential improvement, Raiker says. 

“So, even when we’re thinking about digitalization, there are advantages to just understanding and being able to see how our people and our technology interact,” he says.