IAB203 – Business Process Modelling

Process Discovery: Part 2

Common process discovery methods

Dr. Wasana Bandara

BPM Group
Queensland University of Technology
- Process discovery
- Process discovery challenges
- Common Process discovery methods
Common Process discovery methods

1. Evidence-based
• Document analysis
• Observation
• Automated process discovery

2. Interview-based

3. Workshop-based

Choose one or more on the basis of

context and budget

3
Process discovery methods

1. Evidence-based
• Document analysis
• Observation
• Automated process discovery

2. Interview-based

3. Workshop-based

Choose one or more on the basis of

context and budget

4
 Evidence-Based: Document Analysis
Documents point to existing roles, activities and business objects:
• Process descriptions (ideal scenario)
• Internal policies
• Organization charts
• Employment plans
• Quality certificate reports
• Glossaries and handbooks
• Forms
• Work instructions…

Could be used to gather information before approaching domain experts.

Potential issues:
• May not be process-oriented
• May require abstraction or refinement
• Trustworthiness of the documents

5
Documents: Policy vs Process vs Procedure

What's the Difference?

Process discovery methods

1. Evidence-based
• Document analysis
• Observation
• Automated process discovery

2. Interview-based

3. Workshop-based

Choose one or more on the basis of

context and budget

7
Evidence-Based: Observation
• Follow directly the execution of individual process instances, then
abstract from instance to process level:
• 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

Some advantages:
• Provides a good understanding of the boundaries of the process and its
essential milestones.
• It reveals how a process is conducted in reality today, (which is in
contrast to document analysis that typically captures the past).
• Can get a clear picture on the details (i.e. people involved, data/ systems
input)
• As done in ‘situ’ one may be able to identify other environmental factors
that impact the process efficiency (e.g. layout of the work stations etc.)
8
Evidence-Based: Observation (Cont.)

Potential issues:
• Active role: no big picture
• Passive role: participants’ bias

• it requires access to the people and sights where the process is being worked on
and such access requires the approval of the managers and supervisors of the
corresponding teams
• The issues with people acting differently when they know they are been
observed
• Time consuming
• Can only see some case examples
Process discovery methods

1. Evidence-based
• Document analysis
• Observation
• Automated process discovery

2. Interview-based

3. Workshop-based

Choose one or more on the basis of

context and budget

10
Evidence-Based: Automated Process Discovery
(and process mining)

Discovery
discovered model

event stream

Enhancement
enhanced model
event log

existing model

DB
Conformance /
11
Process Mining: Minimum Data Requirements
• Activity name and timestamp
• Reference to case id

Additional information:
• Activity resource, cost
• Case attributes (e.g. customer reference, type of case…)
Quick Introduction to process discovery through process mining

• https://www.youtube.com/watch?v=-u5y2dST7ng
A detailed introduction to process mining

https://www.youtube.com/watch?v=9yA0H246fTE
Process discovery methods

1. Evidence-based
• Document analysis
• Observation
• Automated process discovery

2. Interview-based

3. Workshop-based

Choose one or more on the basis of

context and budget

15
Interviews
Interview

Validation Modeling

Verification

Approaches:
• forward vs backward
• structured vs unstructured

Assumption: analyst and domain expert share terminology

Pitfall: exceptional behavior tends to be neglected > use questions that aim to
identify such behavior

16
Interview: Show time
 Establish interviewee (process
stakeholder) profile
 Assessing the challenges – never
too early to do so

 Understanding the process

customers’ Environment
 Recap the Understanding

 Analyst’s Inputs on Customer’s

responses

 Assessing Your Solution (if

applicable)
The SCILSS Technique

• Situation
• Concerns
• Issues
• Like to See
• Summary
• Solution

18
Process discovery methods

1. Evidence-based
• Document analysis
• Observation
• Automated process discovery

2. Interview-based

3. Workshop-based

Choose one or more on the basis of

context and budget

19
Workshops
• Gather all key stakeholders together
• Participants interact to create shared understanding
• Typically one process analyst (facilitator), multiple domain
experts, process owner may also attend
• May be software-supported – a model is directly created during
the workshop (typically a separate role – tool operator)
• Model is used as a reference point for discussions
• Alternative: brown-paper workshops
• Usually 3 to 5 half-day sessions

20
 Preparing for the workshops
Set clear scope and purpose
• Selling the workshop concept to stakeholders
• Agree the scope of the information/processes to be captured.

Stakeholder participation
• Ensuring the Participation of the Right Stakeholders.
• Invite the subject matter experts from the business area you are
mapping  - they are essential. It is also preferable to have someone
present who has knowledge of how the process links to other business
areas.
• Keep the number of attendees to ~ 10 people (or less) however -
otherwise the session is likely to become unwieldly and unproductive.
• Get the process owner's buy-in to your project and support for
members of their team to attend your workshop.
Preparing for the workshops
 Logistics
 Decide the timeframe for the workshop
(and then stick to it).
 Try and prevent Murphy’s law
 Includes travel, lighting, and even
“afternoon sugar filled snacks.”

 Warm-up materials
 Send the participants an agenda for the
workshop, with supporting information
which explains why they have been
asked to attend.
 Project-specific information
 Out-of-box thinking preparation

 Give everyone invited an

opportunity to ask questions
before the event and raise any
concerns. Give the same
opportunity to their departmental
heads
Essential utensils to run the workshops
Role of the Facilitator
 Establish professional and
objective tone to the meeting.
 Start and stop the meeting on
time.
 Establish and enforce the “rules”
for the meeting.
 Introduce the goals and agenda for
the meeting.
 Manage the meeting and keep the
team “on track.”
 Facilitate a process of decision and
consensus making, but avoid
participating in the content.
 Make certain that all stakeholders
participate and have their input
heard.
 Control disruptive or unproductive
behavior.
 Different facilitator styles

• The driving engineer style

• The driving artist style

• The catalyzing engineer style

• The catalyzing artist style.

Hjalmarsson et al., (2015)

Dyadic behavioural anchors of the facilitation
styles
• Communication anchor: talks vs listens

• Power anchor: assertive vs non-assertive

• Adaption anchor: static vs flexible

• Disagreement anchor: embraces conflict vs avoids

conflict

• Control anchor: centralized vs decentralized

Hjalmarsson et al., (2015)
 Dyadic behavioural anchors of the facilitation
styles
• Modelling anchor: does model vs lets the group model

• Facilitation anchor: does facilitation vs lets the group

facilitate

• Involvement anchor: involves vs disconnects

• Work approach anchor: structured vs unstructured

• Domain knowledge anchor: domain agnostic vs domain

expert
Hjalmarsson et al., (2015)
Process discovery methods

1. Evidence-based
• Document analysis

r y
• Observation
• Automated process discovery

2. Interview-based
ma
3. Workshop-based
u m
S
28
Discovery methods: strengths and weaknesses
Method Strength Weakness
Document Analysis • Structured information • Outdated material
• Independent from • Wrong level of
availability of abstraction
stakeholders

Observation • Context-rich insight into • Potentially intrusive

process • Stakeholders likely to
behave differently
• Only few cases

Automatic Discovery • Extensive set of cases • Potential issue with data

• Objective data quality and level of
abstraction
Interview • Detailed inquiry into • Requires sparse time of
process process stakeholders
• Several iterations
required before sign-off

Workshop • Direct resolution of • Requires availability of

conflicting views several stakeholders at
the same time

29
Discovery methods: strengths and weaknesses
Summary till now …

31
Dr Wasana Bandara

Business Process Management Group

Information Systems School
Science and Engineering Faculty
Queensland University of Technology
Office Y Block, level 6
Brisbane QLD 4000

E-mail: [email protected]