How To Conduct a Process Mining Project?

Transcription

1 Process Mining: Data Science in Action How To Conduct a Process Mining Project? prof.dr.ir. Wil van der Aalst

2 Drowning in How to conduct a process mining project?

3 Process mining use cases Minimize mean flow time, waiting time, response time, etc. Improve Key Performance Indicators. improve KPIs related to time goal action (improve process) improve KPIs related to costs improve KPIs related to quality use case Maximize service levels, i.e., percentage handled within agreed upon period. (improve control) (handle problem in ad-hoc manner) (,, )

4 Process mining use cases Redesign: Structural Adjust: Non structural (i.e., changes to the Process process mining should temporary) changes, e.g., based on insights, result e.g., in "actionable" adding more resources making the process information. because of fluctuations in more concurrent or case volume. adding controls. improve KPIs related to time goal action (improve process) improve KPIs related to costs Support: Intervene: Systematically Acting using on pre mortem process event mining, insights e.g., for ing related to particular the activity cases most likely to minimize or resources. the flow time. improve KPIs related to quality use case (improve control) (handle problem in ad-hoc manner) (,, )

5 Process mining use cases Identification of bottlenecks to trigger a process that reduces the overall flow time with 30%. Identification of compliance problems using conformance ing. Some of the compliance problems result in ad-hoc interventions whereas others lead to ments of the parameters used for work distribution. Harmonization of two processes after a merger based on a comparison of the actual processes. The goal of such a harmonization is to reduce costs. Predicting of the remaining flow time to improve customer service. Providing ations for resource allocation aiming at a more balanced utilization of workers. Identification of exceptional cases that generate too much additional work. By learning the profile of such cases, they can be handled separately to reduce the overall flow time. Visualization of the 10 most complicated or time consuming cases to identify potential risks. improve KPIs related to time goal action (improve process) (improve control) improve KPIs related to costs use case (handle problem in ad-hoc manner) improve KPIs related to quality (,, )

6 L* lifecycle for process mining

7 L* lifecycle for process mining understanding business understanding Describes the lifecycle of an idealized process mining project (assuming "Lasagna processes"). Stage 2: create and connect current Stage 4: operational

8 Similar to reference describing the lifecycle of a typical mining/bi project "CRISP-DM Process Diagram" by Kenneth Jensen CC BY-SA 3.0 Sample Explore Modify Model Assess CRISP-DM CRoss-Industry Standard Process for Data Mining SEMMA Sample, Explore, Modify, Model and Assess [SAS Institute]

9 Stage 0: Plan and justify Three types of projects: understanding business understanding Stage 2: create and connect -driven ("curiosity" driven) question-driven ("why?") goal-driven (improve KPI) current Plan project. Justify planned activities ("business case"). Stage 4: operational

10 Stage 1: Extract understanding business understanding Stage 2: create and connect Locate, extract and transform event (non-trivial, see previous lectures). Moreover, collect: and other artifacts,, and. current Stage 4: operational Exploit existing (domain) knowledge!

11 Stage 2: Create and connect understanding business understanding Control-flow is the backbone of any process. Stage 2: create control-flow and connect Therefore, first create a suitable well-connected to the available event. current Conformance ing and alignments are key! Stage 4: operational Iterative (like other stages).

12 Stage 3: Create integrated understanding Stage 2: create and connect business understanding Stage 3: create integrated Replay event on to learn about the other perspectives (time,, resources, ). Merge into an overall showing the different perspectives. current Stage 4: operational

13 Stage 4: Operational understanding Stage 2: create and connect business understanding Use current (pre-mortem) for on-the-fly deviation ion, ions, and ations. Only possible for Lasagna processes! current Stage 4: operational

14 Relation to use cases improve KPIs related to time goal action (improve process) understanding business understanding improve KPIs related to costs use case (improve control) (handle problem in ad-hoc manner) improve KPIs related to quality (,, ) Stage 2: create and connect current Stage 4: operational

15 Linking L* to the refined process mining framework understanding business understanding people machines business processes world documents organizations information system(s) provenance s Stage 2: create and connect navigation pre mortem current auditing post mortem cartography de jure de facto current control-flow control-flow Stage 4: operational /rules /rules resources/ organization resources/ organization

16 Linking L* to the refined process mining framework understanding business understanding people machines business processes world documents organizations information system(s) provenance s Stage 2: create and connect navigation pre mortem current auditing post mortem cartography de jure de facto current control-flow control-flow Stage 4: operational /rules /rules resources/ organization resources/ organization

17 Linking L* to the refined process mining framework understanding business understanding people machines business processes world documents organizations information system(s) provenance s Stage 2: create and connect navigation pre mortem current auditing post mortem cartography de jure de facto current control-flow control-flow Stage 4: operational /rules /rules resources/ organization resources/ organization

18 Linking L* to the refined process mining framework understanding business understanding people machines business processes world documents organizations information system(s) provenance s Stage 2: create and connect navigation pre mortem current auditing post mortem cartography de jure de facto current control-flow control-flow Stage 4: operational /rules /rules resources/ organization resources/ organization

19 L* lifecycle for process mining understanding business understanding Stage 2: create and connect current Stage 4: operational

20 Part I: Preliminaries Part III: Beyond Process Discovery Chapter 1 Introduction Chapter 2 Process Modeling and Analysis Chapter 3 Data Mining Chapter 7 Conformance Checking Chapter 8 Mining Additional Perspectives Chapter 9 Operational Support Part II: From Event Logs to Process Models Chapter 4 Getting the Data Chapter 5 Process Discovery: An Introduction Chapter 6 Advanced Process Discovery Techniques Part IV: Putting Process Mining to Work Chapter 10 Tool Support Chapter 11 Analyzing Lasagna Processes Chapter 12 Analyzing Spaghetti Processes Part V: Reflection Chapter 13 Cartography and Navigation Chapter 14 Epilogue