As companies try to extract and make sense of their business processes, they are finding that traditional business process management methodologies (BPM, BI, etc.) are not only expensive and time consuming, but the results, or proposed new processes, can be overly idealistic and disconnected from reality. If companies want to reliably analyze, improve, and manage their processes, they need to
be able to rely on more concrete sources of information and unbiased analysis. In addition, when processes are lengthy or complex, process discovery needs to be automated for accurate results. In this article we’ll be covering what process mining is, how it’s being leveraged for business transformation, and it’s relationship to other key disciplines like process automation and data science.
As part of the broader discipline of process science process mining is an analytical technique that leverages both data mining and model-based process analysis in order to discover or "mine" previously hidden insights about the current state processes through the analysis of event data. An event is a specific instance of an activity and while some of these activities are units of work within a defined business process model, many are not. Professor Wil van der Aalst is called “one of the fathers of process mining”; he realized that analyzing events from IT systems could provide unbiased, detailed insights on a company's processes. He said the following about event data sources:
“Events may take place inside a machine (e.g., an X-ray machine, an ATM, or baggage handling system), inside an enterprise information system (e.g., an order placed by a customer or the submission of a tax declaration), inside a hospital (e.g., the analysis of a blood sample), inside a social network (e.g., exchanging e-mails or twitter messages), inside a transportation system (e.g., checking in, buying a ticket, or passing through a toll booth), etc. Events may be “life events”, “machine events”, or “organization events.”
Collecting Event Data
Since these events can take place pretty much anywhere, event data is gathered from a wide variety of sources. Typical data sources include databases, message logs, ERP systems, CRM systems, BPMS, etc. Event data is further organized, enriched, and filtered into particular views known as event logs which serve as the starting point or “raw material” for process mining. Where data mining starts more broadly with data sets of various kinds, process mining starts specifically with event log data sets but also leverages advanced data science methods and tools (e.g., artificial intelligence, machine learning, etc.) and applies them from a model-driven, process science perspective. This unique approach of process mining as a blend of process science and data science methods is what has led Professor Wil van der Aalst to call process mining “a means to bridge the gap between data science and process science.”
This approach to process mining allows companies to reliably attain an understanding of the events taking place within their business through the extraction of event data. Then this data can be correlated to defined process models.
Automatic Discovery of Business Processes
Once event log data is correlated to business processes, they can be automatically discovered. Process deviations and bottle-necks can be analyzed. Predication models can be generated and animated. New processes can be more accurately documented with less manual effort. And, ultimately, new processes can be automatically generated (ABPM).
“The idea of process mining is to discover, monitor and improve real processes (i.e., not assumed processes) by extracting knowledge from event logs readily available in today’s systems.” -Will Van Der Aalst 
Business Process Mining and Workflow Automation
Another major benefit of sophisticated process mining is its connection to process automation and workflow management. Since process mining helps provide realistic and up-to-date process insights, companies can more effectively target the processes in their organization that would yield the greatest ROI if automated. Gartner’s “Market Guide for Process Mining” states,
"In most cases, tasks are part of processes and operations for which change is the most common characteristic. By accurately assessing the processes to which these tasks belong, we can identify 'hot areas' in the organization where a lot of effort is wasted in repetitive tasks. Then, we can see if these tasks can be partly or fully automated via RPA. This is where process mining can complement RPA perfectly to offer a wider context and help implement this task automation, resulting in generating long term sustainable business value and avoiding the current shortcomings of having a short-term perspective focused on large one-off cost savings"
Complex processes often occur across many departments and people, thus individuals and departments often do not understand them end-to-end. Thus, another benefit of process mining is that we can see the wider context in which tasks occur, and receive performance and conformance analytics. Process mining helps in the initial identification of the tasks that should be automated. Automated tasks can then be better managed once implemented, and quickly fine-tuned as needed. Process mining also enables more accurate calculation of return on investment (ROI).
"Process mining improves the success rate of task-level automation such as RPA through visualizing and understanding the process context, so that when processes change, the automated tasks can adapt and don't lose their relevance. Moreover, process mining helps in spotting and prioritizing opportunities for task-level automation."
By leveraging the major advantages of process mining, namely automated process discovery and accurate performance and conformance analytics, companies can optimize both their process models and performance and fundamentally transform the way they operate and do business.
Be sure to review our use-case focused blogs on intelligent automation:
- Intelligent Automation as a Key Enabler of Tomorrow's Agile Finance Function
- Making the Case for Robotic Process Automation (RPA) in Human Resources
- Why Robotic Process Automation is a Game Changer for the Recruiting Industry
- Process Automation in the Learning Organization: A Deployment Necessity and A Strategic Consideration
- Why ITSM Practitioners Should Understand Robotic Process Automation (RPA)
- How Automation is Transforming Healthcare and Its Potential to Save Lives
 “Data mining can be defined as “the analysis of (often large) data sets to find unsuspected relationships and to summarize the data in novel ways that are both understandable and useful to the data owner” Aalst, Wil van der. Process Mining Data Science in Action. Springer, 2016.