Process Mining: What’s Stopping Us?

Process mining shows great promise in improving process discovery, measurement and auditing, but Australian enterprises have been slow to take it up. This post will explore process mining and discuss some of its benefits, before asking readers to identify what’s holding them back from taking it up. We think it can deliver real and achievable results. Anecdotally, many people seem to feel the same way, but few have taken the step of putting it into practice. We’re interested in your thoughts, so feel free to offer your perspective at the end of this post.

What is process mining?

Many people will already know something about process mining, but for those who don’t, here’s a brief summary. Process mining is a tool-assisted technique for reconstructing business processes from events recorded as transactions in an information system. Logs are ‘mined’ for related patterns of activity which are then (by means of specialised software) reproduced as business process models. In other words, an algorithmic assessment is made of related transactions which is then turned into a diagram showing a sequential set of steps, documented in a formal notation. Because system logs contain authoritative information about system events, a large amount of metadata and measures about processes can be assembled. Much work has been done in research institutions to develop and refine algorithms, and so process mining is generally viewed as one of the more academically-based process management tools.

One of the results of this approach is that process is capable of producing process models that show all the possible paths of a process, including exceptions and escalations. The purpose of process mining is not to produce executable model designs (or ‘to-be’ process models), but to document and analyse existing processes comprehensively and accurately.

What does process mining do?

While process mining is an area of interest for information systems researchers, for practical purposes it can be used to do the following:

• Discovery of business processes, that is, interrogating systems to understand currently undocumented or unknown business processes
• Check the compliance of business processes, that is, auditing to see whether the process as practiced matches a formal prescription for how the process ought to be carried out
• Add extra information to existing process models, such as applying measures and illustrating alternative paths
• Combine and merge separate process models to show how they work in conjunction (or in conflict)
• Analyse process performance.

At its simplest, process mining is an effective way of producing or verifying process models. In other words, it’s a very useful tool for establishing what actually happens in a process, either as discovery or as verification. This understates the value, however, as process mining delivers process models that are correct, complete and contain more information than can be obtained by other means.

Your thoughts

Having briefly recounted what process mining can do we’d like to hear your thoughts by commenting below.