Process Discovery

                       Process Discovery 


Process Discovery  

- The act of gathering information about an existing process and organizing it in terms of an as-is process model. 

- The Process Discovery phase of the BPM lifecycle identifies how to create models that are both correct and complete. 

- This is clearly more than modeling (however, modeling is a part of  it.)

Phases of Process Discovery 

1. Defining the setting. 

2. Gathering information.  

3. Conducting the modeling task. 

4. Assuring process model quality.

Who is involved? 

                           

Domain Expert                                      Process Analyst

Expertise of a Domain Expert 

- Domain Expert has solid knowledge on how a process or  activity is performed. 

- Typically a  domain expert is a process participant. 

- But, it can be the process owner or the manager who closely with  process participants who perform the processes. 

- Suppliers/Customers can also be considered as domain  experts.

Expertise of a Process Analyst 

1. Problem understanding 

               - Episodic knowledge available to get to root of problem 

               - Knowledge organizations helps to structure problem 

2. Problem solving 

               - Trigger identification (problem-related cues) 

               - Hypothesis management (formulation and testing of hypotheses) 

               - Goal setting (what needs to be achieved next) 

               - Top-down strategy driven by analysis goals 

3. Modelling skills 

               - Well-structured and laid out 

               - Systematically labelled 

               - Explicit start and end points of a process 

               - Appropriate granularity and decomposition

Process Discovery Challenges 

1.Fragmented Process Knowledge 

2.Domain Experts think on Instance Level 

3.Knowledge about Process modelling is rare

Challenge 1: Fragmented Process Knowledge 

- A business process defines a set of logically related activities. 

- These activities are assigned to specialized participants. 

- Thus, when collecting knowledge a process analyst needs to  talk to not just one domain experts but different domain experts  who are responsible for different tasks. 

- A domain expert typically has abstract understanding of the  whole process but a very detailed understanding of their own  task. 

Challenge 2:  Domain Experts think on Instance Level 

- Domain experts will easily describe the activities of a  specific case   but they might have problems responding to  general questions. 

- Often said by domain experts: “every case is different”. 

- Process analyst will have to organize and abstract from the  pieces of information provided by the domain experts.

Challenge 3:  Knowledge about Process Modelling is rare

- Domain experts are often not trained to create or read  process models. 

- Thus seeking feedback to a draft process model is  difficult.

Process Discovery Techniques 

- Evidence-based

           Document analysis 

           Observation 

           Process mining 

- Interview-based  

- Workshop-based

     Evidence Based Discovery - Document Analysis

Documents point to existing roles, activities and  business objects. Formal documentation in terms of 

-   Organization chart 

-    Employment plan 

-   Quality certificate report 

-   Internal policies 

-   Glossaries and handbooks 

-   Forms 

-   Work instructions

  Observation 

- Observe what people do at their workplace Active role: play a specific role, 

e.g. customer Passive role: observe participants and their environment 

- Trace business objects in the course of their lifecycle 

- Inspect the work environment.

Automated Discovery via Process Mining 

- Process mining is a process management technique that allows for  the analysis of business processes based on event logs. 

- The basic  idea is to extract knowledge from event logs recorded by an  information system. 

- Process mining aims at improving this by  providing techniques and tools for discovering process, control, data,  organizational, and social structures from event logs. 

- Automatic process discovery makes use of event logs that are stored  by information systems.

Stages of Process Modeling Method 

1.Identify the process boundaries. 

2.Identify activities and events. 

3.Identify resources and their handovers. 

4.Identify the control flow. 

5.Identify additional elements.

1. Identify the process boundaries 

- Under which condition does the process start? (triggers) 

- With which result does it end? (outcomes) 

- Which perspective do you assume? 

- What data are required as input and output to  the  process? 

- Identifying the process boundaries is essential to identifying the process  scope

2. Identify activities and events 

3. Identify resources and their handovers Identify 

4. Identify the control flow

5. Identify Additional Elements 

Extend the process model by capturing artifacts and exception handlers.

- Artifacts – data objects, data stores and their relations to activities  and events via data associations. 

- Exception Handlers – boundary events, exception flows.