Why Data Modeling Is Fundamental

© Copyright 2021 by Peter Aiken Slide # 1
paiken@plusanythingawesome.com+1.804.382.5957 Peter Aiken, PhD
Why Data Modeling
is Fundamental
Peter Aiken, Ph.D.
• I've been doing this a long time
• My work is recognized as useful
• Associate Professor of IS (vcu.edu)
• Institute for Defense Analyses (ida.org)
• DAMA International (dama.org)
• MIT CDO Society (iscdo.org)
• Anything Awesome (plusanythingawesome.com)
• 11 books and dozens of articles
• Experienced w/ 500+ data
management practices worldwide
• Multi-year immersions
– US DoD (DISA/Army/Marines/DLA)
– Nokia
– Deutsche Bank
– Wells Fargo
– Walmart …
© Copyright 2021 by Peter Aiken Slide #
https://plusanythingawesome.com 2

Supporting Your Fundamental Data
Modeling Needs

Data Modeling Challenges
Heterogeneous
Databases
Technologies
Agile Data
Modeling
Semantic
Integration
© 2021 erwin, Inc. All rights reserved. 2
Integrating Modeling
Techniques

Data Modeling Requires Broad DBMS Support and
Integration
Metadata
Transformation
& Exchange
Import
Bridges from
Vendor Tools
Export
Bridges to
Vendor Tools
Data & Object
Modeling
Data
Integration
(EAI, ETL, EII)
Business
Intelligence
(OLAP, Reporting)
Data & Object
Modeling
Data
Integration
(EAI, ETL, EII)
Business
Intelligence
(OLAP, Reporting)

Data Modeling Requires Multiple, Integrated Design
Paradigms
Relational Design
• Structure is driven by how we expect to query the data
• Overarching principle is query optimization – access
all relevant data from fewest containers
• Leads to simple queries with minimal joins: faster =
good candidate for real-time applications
NoSQL Document Design
• Structure is driven by what data we want to capture
and store
• Overarching principle is storage optimization – store
atomic data once
• Leads to complex queries with multiple joins: slower =
poor candidate for real-time applications

Data Vault 2.0
Hubs – A list of unique business keys
Links – An intersection of business keys (two or more)
Satellites – Non-key descriptive columns that change over time
Bridges – A combination of primary and business keys spread across multiple Hubs and Links
PIT – “Point in Time” combination of primary and business keys from a single Hub and its surrounding Satellites
Reference – A collection of code and description lookup structures that are generally resolved as run-time queries
Data Vault is a comprehensive methodology (rules, best
practices, standards, process designs and more)
composed of three pillars: Architecture, Model, and
Methodology. Data Vault is designed and used for solving
enterprise level issues such as: agility, scalability,
flexibility, auditability and consistency.”
The Data Vault Alliance
“

User Access and Permission
Management
Model Check In/Check Out
Model Change Control with Version
Management
Concurrent Modeling with Conflict
Resolution
Cross Model Reporting
Centralized Standards Access and
Management
Model Mart
Data
Analyst
Data
Architect
DBA &
Developer
Model Management Services
Model
Management
Services
Governed Collaboration for Data Modeling

Semantic Integration – Microsoft Common Data Model
(CDM)
Automatically transform the Microsoft CDM into
a graphical model, complete with business data
constructs and semantic metadata
Feed existing data models and database
designs with reusable CDM constructs and
semantics
Manage ongoing integration and reuse of CDM
best practices through compare, synchronization
and automated model templates

Optimize Data Design, Dev/OPS, Literacy &
Governance with Data Modeling
8
Data
Modeling
Lower
Design
Costs
Reduced
Design
Risks
Improved
Design
Quality
Increased
Design
Agility
REDUCE TOTAL COST OF OWNERSHIP
ACCELERATE TIME TO VALUE
Assuring Business Alignment
Reducing Expensive Re-Work
Easing Integration
Enabling Collaboration
© 2021 erwin, Inc. All rights reserved.

https://plusanythingawesome.com
Why Data Modeling
is Fundamental
Program
• Data Management Contextual Overview
• Motivation
– of systems/components
– Data is not well understood
• Why data modeling & what is it?
– Model represents our understanding of the
– Fundamental, foundational system characteristics
– Shared between systems and humans
• Fundamentals
– The power of the purpose statement
– Understanding data centric thinking
– Data modeling compliments other architecture/engineering techniques
– Challenges beyond data modeling
• Take Aways, References, Q&A
X
2020 AA market value ~ $6b
AAdvantage valued between $19.5-$31.5
United market value ~ 9$b
MileagePlus ~ $22b
https://www.forbes.com/sites/advisor/2020/07/15/how-airlines-make-billions-from-monetizing-frequent-flyer-programs/?sh=66da87a614e9
Data
Assets
Financial
Assets
Real
Estate Assets
Inventory
Assets
Non-
depletable
Available for
subsequent
use
Can be
used up
Can be
used up
Non-
degrading √ √ Can degrade
over time
Can degrade
over time
Durable Non-taxed √ √
Strategic
Asset √ √ √ √
Data Assets Win!
• Today, data is the most powerful, yet underutilized and poorly managed
organizational asset
• Data is your
– Sole
– Non-depletable
– Non-degrading
– Durable
– Strategic
• Asset
– Data is the new oil!
– Data is the new (s)oil!
– Data is the new bacon!
• As such, data deserves:
– It's own strategy
– Attention on par with similar
organizational assets
– Professional ministration
to make up for past neglect
Asset: A resource controlled by the organization as a result of past events or
transactions and from which future economic benefits are expected to flow [Wikipedia]
Data Assets Win!

Why Model?
• Would you build a house without an
architecture sketch?
• Model is the sketch of the system to be
built in a project.
• Would you like to have an estimate how
much your new house is going to cost?
• Your model gives you a very good idea of
how demanding the implementation work
is going to be!
• If you hired a set of constructors from all
over the world to build your house, would
you like them to have a common
language?
• Model is the common language for the
project team.
• Would you like to verify the proposals of
the construction team before the work gets
started?
• Models can be reviewed before thousands
of hours of implementation work will be
done.
• If it was a great house, would you like to
build something rather similar again, in
another place?
• It is possible to implement the system to
various platforms using the same model.
• Would you drill into a wall of your house
without a map of the plumbing and electric
lines?
• Models document the system built in a
project. This makes life easier for the
support and maintenance!
powerpivotpro.com
Augusta Ada King (aka Lady Ada, Countess of Lovelace)
https://people.well.com/user/adatoole/bio.htm
Jacquard machine 1804
≈
• 8,000+ years
• formalize
practices
• GAAP

Unrefined
data management
definition
Sources
Uses
Data Management
More refined
data management
definition
Sources
Reuse
Data Management
➜ ➜

Better still data management definition
Data Governance
Data Assets/Ethical Framework
Sources
➜ Use
➜Reuse
➜
You can accomplish
Advanced Data Practices
without becoming proficient
in the Foundational Data
Practices however
this will:
• Take longer
• Cost more
• Deliver less
• Present
greater
risk
(with thanks to
Tom DeMarco)
Data Management Practices Hierarchy
Advanced
Data
Practices
• MDM
• Mining
• Big Data
• Analytics
• Warehousing
• SOA
Foundational Data Practices
Data Platform/Architecture
Data Governance Data Quality
Data Operations
Data Management Strategy
T
e
c
h
n
o
l
o
g
i
e
s
C
a
p
a
b
i
l
i
t
i
e
s

Digital Insight
• Subtract data from digital
and what do you have?
• Subtract digital from data
and you still have data
https://www.linkedin.com/in/mark-johnson-518a752/
DIGITAL
DATA
?
DIGITAL
DATA
DATA
Recent Technology Realization
12
Recent

( Bad Data ) + Anything Awesome ( will always yield ) Bad Results
Garbage In ➜ Garbage Out!
+
Perfect
Model
Garbage
Data
Garbage
Results
Data
Warehouse
Machine
Learning
Business
Intelligence
Block Chain
AI
MDM
Data
Governance
Analytics
Technology
GI➜GO!

Perfect
Model
Garbage
Data
Garbage
Results
Data
Warehouse
Machine
Learning
Block Chain
AI
MDM
Analytics
Technology
Data
Governance
GI➜GO!
Business
Intelligence
Perfect
Model
Quality
Data
Good
Results
Data
Warehouse
Machine
Learning
Business
Intelligence
Block Chain
AI
MDM
Analytics
Technology
Data
Governance
Quality In ➜ Quality Out!

It isn't possible to go digital
Digital
a
By just spelling 'data'
Dat

It requires more work
Data
a
19
Metadata
Management
20
Data
Management
Body
of
Knowledge
(DM
BoK
V2)
Practice
Areas
from The DAMA Guide to the Data Management Body of Knowledge 2E © 2017 by DAMA International
• Analysis
• Database Design
• Implementation
• Additional data
development

DAMA DM BoK: Data Development
from The DAMA Guide to the Data Management Body of Knowledge © 2009 by DAMA International
Why Data Modeling
is Fundamental
Program
• Motivation
• Fundamentals
X

Data Architectures: here, whether you like it or not
deviantart.com
• All organizations
have data
architectures
– Some are better
understood and
documented (and
therefore more
useful to the
organization) than
others
Levels of Abstraction, Completeness and Utility
• Models more downward facing - detail
• Architecture is higher level of abstraction -
integration
• In the past architecture attempted to gain
complete (perfect) understanding
– Not timely
– Not feasible
• Focus instead on
architectural components
– Governed by a framework
– More immediate utility
• http://www.architecturalcomponentsinc.com

How are components expressed as architectures?
• Details are
organized into
larger components
• Larger components
are organized into
models
• Models are
organized into
architectures
(composed of
architectural
components)
A B
C D
A B
C D
A
D
C
B
Intricate
Dependencies
Purposefulness
How are data structures expressed as architectures?
• Attributes are organized into entities/objects
– Attributes are characteristics of "things"
– Entitles/objects are "things" whose
information is managed in support of strategy
– Example(s)
• Entities/objects are organized into models
– Combinations of attributes and entities are
structured to represent information requirements
– Poorly structured data, constrains organizational
information delivery capabilities
– Example(s)
• Models are organized into architectures
– When building new systems, architectures are used to plan development
– More often, data managers do not know what existing architectures are and -
therefore - cannot make use of them in support of strategy implementation
Intricate
Dependencies
Purposefulness
THING
Thing.Id #
Thing.Description
Thing.Status
Thing.Sex.To.Be.Assigned
Thing.Reserve.Reason

Q: What is an Attribute?
• What does the existence of this attribute tell us?
– Clubs need to be identified (#) separately from one another
– Club-specific information is likely maintained
– Some concept (organization) exists above the 'club level'
– ...
A: Attribute Definition
• Attributes describe an entity and attribute values describe
“instances of business things”

Entities organized into a model
• Defines mandatory/optional relationships using minimum/
maximum occurrences from one entity to another
Data architectures are composed of data models

Working While Bleeding
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$$$$$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$$$$$ $
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$$$$$
$
$
$
$
$
Data incoherence is a hidden expense
• How does maltreated data cost money?
• Consider the opposite question:
– Were your systems explicitly designed to
be integrated or otherwise work together?
– If not then what is the likelihood that they
will work well together?
• Organizations spend 20-40% of their IT
budget evolving their data - including:
– Data migration
• Changing the location from one place to another
– Data conversion
• Changing data into another form, state, or product
– Data improving
• "Inspecting and manipulating, or re-keying data to prepare it for
subsequent use" - Source: John Zachman
PETER AIKEN WITH JUANITA BILLINGS
FOREWORD BY JOHN BOTTEGA
MONETIZING
DATA MANAGEMENT
Unlocking the Value in Your Organization’s
Most Important Asset.

As a topic, data is ...
Complex &
detailed
• Outsiders do not
want to hear about
or discuss any
aspects of
challenges/solutions
• Most are unqualified
re: architecture/
engineering
Taught
inconsistently
• Focus is on
technology
• Business impact is
not addressed
Not well
understood
• (Re)learned by
every
workgroup
• Lack of standards/
poor literacy/
unknown
dependencies
Wally Easton Playing Piano
https://www.youtube.com/watch?v=NNbPxSvII-Q
33
Bad Data Decisions Spiral
Bad data decisions
Technical deci-
sion makers are not
data knowledgable
Business decision
makers are not
data knowledgable
Poor organizational outcomes
Poor treatment of
organizational data
assets
Poor
quality
data

Why Data Modeling
is Fundamental
Program
• Motivation
• Fundamentals
X
Data Modeling Definition
• Modeling = Analysis and design
method used to
– Define and analyze data requirements
– Design data structures that support
these requirements
• Model = set of data specifications
and related diagrams that reflect
requirements and designs
– Representation of something in our
environment
– Employs standardized text/symbols to
represent data attributes (grouped into
data elements) and the relationships
among them
– Integrated collection of specifications
and related diagrams that represent
data requirements and design

Data Modeling
• Modeling = complex process involving interaction between people
and with technology that don’t compromise the integrity or security
of the data
– Good data models accurately
express and effectively communicate
data requirements and
quality solution design
• Modeling approach
(guided by 2 formulas):
– Purpose + audience = deliverables
– Deliverables + resources + time = approach
Data Models Facilitate
• Formalization
– Data model documents a single,
precise definition of data requirements
and data-related business rules
• Communication
– Data model is a bridge to understanding data
between people with different levels and types of experience.
– Helps understand business area, existing application, or impact of modifying an
existing structure
– May also facilitate training new business and/or technical staff
• Scope
– Data model can help explain the data concept and scope of purchased
application packages

ANSI-SPARC 3-Layer Schema
1. CONCEPTUAL - Allows independent
customized user views:
– Each should be able to access the same
data, but have a different customized
view of the data.
2. LOGICAL - This hides the physical
storage details from users:
– Users should not have to deal with
physical database storage details. They
should be allowed to work with the data
itself, without concern for how it is
physically stored.
3. PHYSICAL - The database
administrator should be able to
change the database storage
structures without affecting the
users’ views:
– Changes to the structure of an
organization's data will be required. The
internal structure of the database should
be unaffected by changes to the physical
aspects of the storage.
For example, a changeover to a new DBMS
technology. The database administrator
should be able to change the conceptual or
global structure of the database without
affecting the users.
Families of Modeling Notation Variants
Information Engineering

What is a Relationship?
• Natural associations between two or more entities
Ordinality & Cardinality
• Defines mandatory/optional relationships using minimum/
maximum occurrences from one entity to another
A BED is placed
in one and only
one ROOM A ROOM
contains zero
or more BEDS
A BED is occupied by zero or more
PATIENTS
A PATIENT
occupies at least
one or more BEDS
ROOM
BED
PATIENT

Q: What is the proper relationship for these entities?
Eventually One or Many (optional)
Eventually One (optional)
Zero, or Many (optional)
One or Many (mandatory)
Exactly One (mandatory)
Possible Entity Relationship Cardinality Options

informed information investing over technology acquisition activities
Person Job Class
Position
BR1) One EMPLOYEE
can be associated with one
PERSON
BR2) One EMPLOYEE can be
associated with one POSITION
Manual
Job Sharing
Manual
Moon Lighting
Employee
Person Job Class
Employee Position
BR1) Zero, one, or more
EMPLOYEES can be associated
with one PERSON
BR2) Zero, one, or more EMPLOYEES
can be associated with one POSITION
Job Sharing
Moon Lighting

Data structures must be specified prior to IT development/acquisition
(Requires 2 structural loops more
than the more flexible data structure)
More flexible data structure Less flexible data structure
Understanding
• Definition:
– 'Understanding an architecture'
– Documented and articulated as a digital blueprint
illustrating the
commonalities and
interconnections
among the
architectural
components
– Ideally the understanding
is shared by systems and humans

Data Modeling Process
1. Identify
entities
2. Identify key
for each entity
3. Draw rough
draft of entity
relationship
data model
4. Identify data
attributes
5. Map data
attributes to
entities
Model evolution is good, at first ...
1. Identify
entities
2. Identify key
for each entity
3. Draw rough
draft of entity
relationship
data model
4. Identify data
attributes
5. Map data
attributes to
entities

Relative use of time allocated to tasks during modeling
Preliminary Modeling Wrapup
Activity activities Cycles activities
Evidence Analysis
collection &
analysis Collection
Project
coordination
requirements Declining coordination requirements
Target
system
analysis Increasing amounts of target system analysis
Modeling Validation
cycle
focus Refinement
Don’t Tell Them That You Are Modeling!
Then make some appropriate
connections between your objects
Just write some stuff down
Then arrange it

Table Handling
• A table is a collection of data items that have the
same description, such as account totals or monthly averages; it
consists of a table name and subordinate items called table
elements.
– Under representation of other database characteristics causes confusion and
introduces risk to organizational data capabilities
• In this example, the table consists of
Song, Album
• and length?
• No, iTunes uses
– Song Start Time
– Song Stop Time
• More flexible
and less risk
https://www.ibm.com/support/knowledgecenter/en/SS6SG3_4.2.0/com.ibm.entcobol.doc_4.2/PGandLR/tasks/tptbl02.htm
There are correct ways to organize data
• Optimization can be done for:
– Flexibility
– Adaptability
– Retrievability
– Risk reduction
– ...
• Techniques include:
– Data integrity
– Smart codes bad/dumb codes good
– Architecture (table joins)
– ...

(a hypothetical portion of the) iTunes database
• What information is lost if we delete record #1?
Record Purchaser ID Song Price
1 Peter We Met Today $0.99
2 Peter My Mother's Voice $1.29
3 Peter Fortune Smiles $0.99
4 Lolly Thousand Pieces of Gold $0.99
(a hypothetical portion of the) iTunes database: Deletion Anomaly
• Question:
– What information is lost if we delete record #1?
• Answer:
– We loose the fact that Peter purchased "We Met Today"
– We loose the fact that "We Met Today" costs $0.99
– This is usually undesirable and unintended
Row Purchaser ID Song Price
1 Peter We Met Today $0.99

Student Activities File: Insertion Anomalies
• Question:
– Suppose we want to add new song SCUBA and that it costs $1.29?
• Answer:
– Cannot enter it until a purchaser buys SCUBA
– We cannot insert a full row until we have an additional fact about that row
5 ??? SCUBA $1.29
Student Activities File: Update Anomalies
• Question:
– Suppose we want to increase the price of 'We Met Today'
from $0.99 to $1.29?
• Answer:
– Change to data items such as Song requires examination of every single record
– Will not catch spelling errors - such as "We met Toddy"
1 Peter We Met Toddy $0.99
5 Lolly SCUBA $1.29

How Should it be Done? (In General)
• As much as possible,
store 1 fact per row
– Row 5 is a good example
as it shows both that
purchaser Lolly has
purchased SCUBA and
that SCUBA costs $0.99
– These are two distinct facts and are correctly stored in two tables sharing a
formal relationship
– More remains codes
1 Peter We Met Toddy $0.99
5 Lolly SCUBA $0.99
PRICING
Row Song Price
1 We Met Today $1.29
2 My Mother's Voice $1.29
3 Fortune Smiles $0.99
4 Thousand Pieces of
Gold
$0.99
5 SCUBA $0.99
PURCHASES
Row Purchaser ID Song
1 Peter We Met Toddy
2 Peter My Mother's Voice
3 Peter Fortune Smiles
4 Lolly Thousand Pieces of Gold
5 Lolly SCUBA
6 Pat SCUBA
How Should it be Done? (Joining Tables)
• Data from the two tables is joined to provide requested information
• Purchaser PETER is now properly registered to own "We Meet
Today"
• The price change for SCUBA has been resolved
• PRICING table is a better engineered solution
PRICING
Row Song Price
Gold
$0.99
5 SCUBA $1.29
PURCHASES
1 Peter We Met Today
5 Lolly SCUBA
6 Pat SCUBA
60
(each price instance can provide context for many purchases)

How Should it be Done? (Connection Types)
• Defines mandatory/
optional relationships
using minimum/
maximum
occurrences from
one entity to another
PRICING
Row Song Price
Gold
$0.99
5 SCUBA $1.29
PURCHASES
1 Peter We Met Toddy
5 Lolly SCUBA
6 Pat SCUBA
How Should it be Done? (Smart codes bad, dumb codes good)
• 804 → N zero N → long distance call signaling
– All telephone switching equipment (hardware) had to
be changed to not route calls to long distance if the
hardware 'saw' a zero in the middle of a 3 digit number
• Course listings
– "You can not add another
undergraduate business
computer course"
• A large organization has to
expand a primary master data item by a number of digits
– Requires upwards of 100,000 changes to be managed
BUS360
Business
Computer
Courses
BUS361
BUS362
BUS363
BUS364
BUS365
BUS366
BUS367
BUS368
BUS369
BUS3??
https://www.youtube.com/watch?v=_f1gwAGfZs0&frags=pl%2Cwn

Why Data Modeling
is Fundamental
Program
• Motivation
• Fundamentals
X
!

!

!

!

64
Organizational Needs
become instantiated
and integrated into a
Data Models
Informa(on)System)
Requirements
authorizes and
articulates
satisfy
specific
organizational
needs
The process is iterative
Data models and data architectures are
developed in response to needs

Bed
Entity: BED
Purpose: This is a substructure within the room
substructure of the facility location. It
contains information about beds within rooms.
Attributes: Bed.Description
Bed.Status
Bed.Sex.To.Be.Assigned
Bed.Reserve.Reason
Associations: >0-+ Room
Status: Validated
Keep them focused on data model purpose
• The reason we are locked in
this room is to:
– Mission: Understand formal
relationship between soda and
customer
• Outcome: Walk out the door with a data
model this relationship
– Mission: Understand the
characteristics that differ between
our hospital beds
• Outcome: We will walk out the door when
we identify the top three traits that
represent the brand.
– Mission: Could our systems
handle the following business rule
tomorrow?
– "Is job-sharing permitted?"
• Outcomes: Confirm that it is possible to
staff a position with multiple employees
effective tomorrow
selects and pays for
given to
Soda
Customer
selects
can be filled by zero or 1
Employee Position
has exactly 1
How does our
perspective change:
the primary means of
tracking a patient
Standard definition reporting does not provide conceptual context
BED
Something you sleep in

Entity: BED
Data Asset Type: Principal Data Entity
Purpose: This is a substructure within the room
substructure of the facility location. It contains
information about beds within rooms.
Source: Maintenance Manual for File and Table
Data (Software Version 3.0, Release 3.1)
Attributes: Bed.Description
Bed.Status
Bed.Sex.To.Be.Assigned
Bed.Reserve.Reason
Associations: >0-+ Room
Status: Validated
The Power of the Purpose Statement
IT Project or Application-Centric Development
https://plusanythingawesome.com Original articulation from Doug Bagley @ Walmart 68
Data/
Information
IT
Projects
• In support of strategy, organizations
implement IT projects
• Data/information are typically considered
within the scope of IT projects
• Problems with this approach:
– Ensures data is formed to the applications and
not around the organizational-wide information
requirements
– Process are narrowly formed around
applications
– Very little data reuse is possible
Strategy

Data-Centric Development
https://plusanythingawesome.com Original articulation from Doug Bagley @ Walmart 69
Data/
Information
IT
Projects
• In support of strategy, the organization
develops specific, shared data-based
goals/objectives
• These organizational data goals/
objectives drive the development of
specific IT projects with an eye to
organization-wide usage
• Advantages of this approach:
- Data/information assets are developed from an
organization-wide perspective
- Systems support organizational data
needs and compliment organizational
process flows
- Maximum data/information reuse
Strategy
the Data Doctrine® (V2)
We are uncovering better ways of developing
IT systems by doing it and helping others do it.
Through this work we have come to value:
data programs driving IT programs
stable, shared organizational data over IT component evolution
data reuse over the acquisition of new data sources
That is, while there is value in the items on
the right, we value the items on the left more.
Source: theagiledoctrine.org

Typically Managed Architectures
• Business Architecture
– Goals, strategies, roles, organizational structure, location(s)
• Process Architecture
– Arrangement of inputs -> transformations = value -> outputs
– Typical elements: Functions, activities, workflow, events, cycles, products, procedures
• Systems Architecture
– Applications, software components, interfaces, projects
• Security Architecture
– Arrangement of security controls relation to IT Architecture
• Technical Architecture/Tarchitecture
– Relation of software capabilities/technology stack
– Structure of the technology infrastructure of an enterprise, solution or system
– Typical elements: Networks, hardware, software platforms, standards/protocols
• Data / Information Architecture
– Arrangement of data assets supporting organizational strategy
– Typical elements: specifications expressed as entities, relationships, attributes,
definitions, values, vocabularies
1 in 10 organizations manage 1
or more of these formally
Data Modeling Example #1
from The DAMA Guide to the Data Management
Body of Knowledge © 2009 by DAMA International
Primary
deliverables
become reference
material
Model Purpose Statement:
This model codifies the official
vocabulary to be used when
describing aspects of any of the
following organizational concepts:
– Subscriber
– Account
– Charge
– Bill

fuel
rent-rate
phone-rate
phone-call
rental
agreement
customer
auto
repair
history
phone-unit
Source: Chikofsky 1990
– fuel
– customer
– auto
– rental agreement
– rent-rate
– phone-call
– phone-rate
– phone-unit
– repair history
It is documentation shown
during the on-
boarding process
Interpretations:
1. Car rental company
2. Rental agreement is central
3. No direct connection between
customer and contract
4. Contract must have a
customer
5. Nothing structural prevents
autos from being rented to
multiple customers
6. Phone units are tied to rentals
salesperson
name
commission
rate
invoice # amount date paid
customer
name
address
customer #
date
order #
price
quantity
order #
item #
quantity
on hand
description
supplier
item # cost
SALESPERSON
INVOICE
ORDER
CATALOG
LINE ITEM
74
• Sales commission-based pricing information
• Difficult to change a customer address
• Price not included in the catalog
• Easy to implement variable pricing - difficult to implement
standard pricing - is standard pricing implemented
• Sales person information is not directly tied to the order
• Do sales people sell things that are shipped quickly so they get
their commission quicker?
• Nothing prohibits a sales from having multiple
sales persons
• Multiple invoices are allowed for a single order
• Partial shipment is allowed
• Data base cannot tell what part of an order the
invoice pertains to
vocabulary and specific
operational rules to be used when
– salesperson
– invoice
– order
– line item
– catalog

DISPOSITION Data Map
describing disposition related organizational concepts:
– user
– admission
– discharge
– encounter
– facility
– provider
– diagnosis
75
Data Model #4: DISPOSITION
• At least one but possibly more system USERS enter the
DISPOSITION facts into the system.
• An ADMISSION is associated with one and only one
DISCHARGE.
• An ADMISSION is associated with zero or more
FACILITIES.
• An ADMISSION is associated with zero or more
PROVIDERS.
• An ADMISSION is associated with one or more
ENCOUNTERS.
• An ENCOUNTER may be recorded by a system USER.
• An ENCOUNTER may be associated with a PROVIDER.
• An ENCOUNTER may be associated with one or more
DIAGNOSES.
76
ADMISSION Contains information about patient admission
history related to one or more inpatient episodes
DIAGNOSIS Contains the International Disease Classification
(IDC) of code representation and/or description
of a patient's health related to an inpatient code
DISCHARGE A table of codes describing disposition types
available for an inpatient at a FACILITY
ENCOUNTER Tracking information related to inpatient
episodes
FACILITY File containing a list of all facilities in regional
health care system
PROVIDER Full name of a member of the FACILITY team
providing services to the patient
USER Any user with access to create, read, update,
and delete DISPOSITION data
episodes
health care system
episodes
health care system
Death must be a disposition code!

As Is Information
Requirements
Assets
As Is Data Design Assets As Is Data Implementation
Assets
Existing
New
Modeling in Various Contexts
O2 Recreate
Data Design
Reverse Engineering
Forward engineering
O5 Reconstitute
Requirements
O9
Reimplement
Data
To Be Data
Implementation
Assets
O8
Redesign
Data
O4
Recon-
stitute
Data
Design
O3 Recreate
Requirements
O6
Redesign
Data
To Be
Design
Assets
O7 Re-
develop
Require-
ments
To Be
Requirements
Assets
O1 Recreate Data
Implementation
Metadata
77
Modeling Options
O-1 data implementation (e.g., by recreating descriptions of implemented file
layouts);
O-2 data designs (e.g., by recreating the logical system design layouts); or
O-3 information requirements (e.g., by recreating existing system specifications and
business rules).
O-4 data design assets by examining the existing data implementation (when
appropriate O-1 can facilitate O-4); and
O-5 system information requirements by reverse engineering the data design O-4.
(Note: if the data design doesn't exist O-4 must precede O-5.)
O-6 transforming as is data design assets, yielding improved to be data designs that
are based on reconstituted data design assets produced by O-2 or O-4 and
(possibly O-1);
O-7 transforming as is system requirements into to be system requirements that are
based on reconstituted system requirements produced by O-3 or O-5 and
(possibly O-2);
O-8 redesigning to be data design assets using the to be system requirements
based on reconstituted system requirements produced by O-7; and
O-9 re-implementing system data based on data redesigns produced by O-6 or O-8.

Model Evolution Framework
Conceptual Logical Physical
Validated
Not Validated
Every modeling change can be mapped
to a transformation in this framework!
Model Evolution (better explanation)
As-is To-be
Technology
Independent/
Logical
Technology
Dependent/
Physical
abstraction
Other logical
as-is data
architecture
components

Pick any two!
and there are
still tradeoffs
to be made!
Data Models Used to Support Strategy
• Flexible, adaptable data structures
• Cleaner, less complex code
• Ensure strategy effectiveness measurement
• Build in future capabilities
• Form/assess merger and acquisitions strategies
Employee
Type
Employee
Sales
Person
Manager
Manager
Type
Staff
Manager
Line
Manager
Adapted from Clive Finkelstein Information Engineering Strategic Systems Development 1992

How do Data Models
Support Organizational Strategy
• Consider the opposite question:
– Were your systems explicitly designed to
be integrated or otherwise work together?
– If not then what is the likelihood that they
will work well together?
– In all likelihood your organization is
spending between much of its IT budget
compensating for poor data structure integration
– They cannot be helpful as long as
their structure is unknown
• Two answers
– Achieving efficiency and effectiveness goals
– Providing organizational dexterity for
rapid implementation
Typical focus of a
database modeling effort
Data Modeling Ensures Interoperability
Program F
Program E
Program D
Program G
Program H
Application
domain 2
Application
domain 3
Program I
Typical focus of a
software engineering effort
Program A

Typical focus of a
database modeling effort
Data Models Ensure Interoperability
Program F
Program E
Program D
Program G
Program H
Application
domain 2
Application
domain 3
Program I
Typical focus of a
software engineering effort
Program A
D
a
t
a
M
o
d
e
l
D
a
t
a
M
o
d
e
l
D
a
t
a
M
o
d
e
l
D
a
t
a
M
o
d
e
l
D
a
t
a
M
o
d
e
l
D
a
t
a
M
o
d
e
l
Program F
Program E
Program D
Program G
Program H
Program I
Application
domain 2
Application
domain 3
D
a
t
a
M
o
d
e
l
D
a
t
a
M
o
d
e
l
D
a
t
a
M
o
d
e
l
Data Model Focus has Great Potential Business Value
• How are decisions
about the range and
scope of common data
usage, made?
• Analysis scope is on
use of data to support a
process
• Problems caused by
data exchange or
interface problems
• Goals often connect
strategic and
operational
• One data model is ideal
D
a
t
a
M
o
d
e
l
Program A

Why Data Modeling
is Fundamental
Program
• Motivation
• Fundamentals
X
Event Pricing
• 20% off
directly from the publisher on
select titles
• My Book Store @
http://plusanythingawesome.com
• Enter the code
"anythingawesome" at the
Technics bookstore checkout
where it says to
"Apply Coupon"
anythingawesome

Use Models to
• Store and formalize information
• Filter out extraneous detail
• Define an essential set of
information
• Help understand complex system behavior
• Gain information from the process of
developing and interacting with the model
• Evaluate various scenarios or other
outcomes indicated by the model
• Monitor and predict system responses
to changing environmental conditions
https://www.youtube.com/watch?v=J5YR0uqPAI8
Data Modeling for Business Value
• Goal must be shared IT/business understanding
– No disagreements = insufficient communication
• Data sharing/exchange is automated and
dependent on successful engineering/architecture
– Requires a sound foundation of data modeling basics
(the essence) on which to build technologies
• Modeling characteristics evolve during the analysis
– Different model instances may be useful to different analytical problems
• Incorporate motivation (purpose statements) in all modeling
– Modeling is a problem defining as well as a problem solving activity
• Use of modeling is more important than selection of a specific
method
• Models are often living documents
• Models need to be available in an easily searchable manner
• Utility is paramount
– Adding color and diagramming objects customizes models and allows for a more
engaging and enjoyable user review process
Inspired by: Karen Lopez http://www.information-management.com/newsletters/enterprise_architecture_data_model_ERP_BI-10020246-1.html?pg=2

Upcoming Events
Business Value through
Reference & Master Data
Strategies
13 July 2021
Getting (Re)Started with
Data Stewardship
10 August 2021
Approaching Data Quality
Engineering
14 September 2021
Brought to you by:
Time: 19:00 UTC (2:00 PM NYC) | Presented by: Peter Aiken, PhD
paiken@plusanythingawesome.com +1.804.382.5957
Questions?
Thank You!
Book a call with Peter to discuss anything - https://plusanythingawesome.com/OfficeHours.html
+ =

Why Data Modeling Is Fundamental

More Related Content

What's hot (20)

Similar to Why Data Modeling Is Fundamental (20)

More from DATAVERSITY (20)

Recently uploaded (20)

Why Data Modeling Is Fundamental