What's Next in Process Mining? Context. It's Everything
Context is king, they say. And rightfully so.
Whatever you do, knowing the circumstances and conditions in which you do it, is the best way to navigate the never-ending labyrinth of ideas. And then you have the ability to take the proper course of action.
The father of Process Mining, prof. Wil van der Aalst, thinks so too.
In his publication (written in cooperation with Schahram Dustdar from the Technical University of Vienna) “Process Mining Put Into Context”, he says that analyzed cases always happen under certain context.
“Existing process mining techniques tend to use a rather narrow context, i.e., only the instance in isolation is considered,” the document reads. “However, the handling of cases is influenced by a much broader context. Therefore, analysis should not abstract from anything not directly related to the individual instance.”
The data from a Process Mining analysis needs to be enhanced and enriched by the context in which they happen. What does it mean in reality? Let's see some examples.
Deeper Insight on Employees' Performance
Do you remember John's case from our previous article on why Process Mining is not a recommendation machine?
Here it is in brief: Process Mining analysis showed you that John might be a potential bottleneck. But thanks to talking to the process owner in your company, you found out that there are other issues, not that John's a bad employee.
An ideal case of Process Mining put into context. How?
In your systems, you have all the information about John - his seniority, experience, education, trainings, all the history. But the analysis could not cross-examine this abundance of information, the context was missing.
And as a result, you might potentially come to the conclusion that you need to dismiss or exchange a well-performing employee.
With Process Mining put into context, once it is able to pool together contextual information, you'll get a full picture of what is going on with your processes, and you'll be able to make better and more informed decisions.
Recording the User Interface
When running an RPA initiative, one of the things you need to do is the recording of the screen, so that the bot will know what activities to perform and how to do it.
However, during the recording, the employee that is being recorded, did not only use the applications needed for that specific task.
They also open emails, check their Facebook, search on Google. Therefore, after the recording, an analyst or consultant needs to say which systems are not of interest for the RPA bot, such as Explorer or Outlook.
With proper contextual data, Process Mining analysis will only take in information from the systems that the users engage with when performing the task.
If we looked at the broader perspective of Process Mining context, we'll end up in context within industries. Otherwise known as industrial benchmarking.
Imagine how valuable the Process Mining analysis would be once you had data on performance of the OTC or PTP processes from other companies within your industry.
There is a lot of experience floating around, and if you have this data, you can use smart algorithms that will say “This tends to be a bottleneck within this process in your industry, there is a potential problem, watch out for that.”
However, at the moment, the industry's benchmarking challenge is how to anonymously share such “personal” data as the performance of a specific process is. Once the time comes, the benefits both for individual companies as well as for industries as such, will be remarkable.
Context In Your Company
Adding context to Process Mining analysis is a pragmatic step towards running a transparent organization that can rely on its own processes. Only then will it become a competitive business open for what the future will bring to the table.
When analyzing processes in your company, where and how do you look for context? Who is your contact person?
Photo by Paul Bence on Unsplash
Michal Tomek Writer & Content Manager at Minit
17. 09. 2019