Data modelling 101

Data Modelling 101
C H R I S T O P H E R B R A D L E Y
D A T A M A N A G E M E N T A D V I S O R S

Christopher Bradley
I N F O R M AT I O N M A N A G E M E N T S T R AT E G I S T
Blog: Information Management, Life & Petrol
http://infomanagementlifeandpetrol.blogspot.com
@InfoRacer
LI: uk.linkedin.com/in/christophermichaelbradley/
T: +44 7973 184475
E: Chris.Bradley@DMAdvisors.co.uk

Christopher Bradley
Chris has 35 years of Information Management
experience & is a leading Independent Information
Management strategy advisor.
In the Information Management field, Chris works with
prominent organizations including Vodafone, BT, HSBC,
Celgene, GSK, Pfizer, Icon, Quintiles, Total, Barclays, ANZ,
GSK, Shell, BP, Statoil, Riyad Bank & Aramco. He
addresses challenges faced by large organisations in
the areas of Data Governance, Master Data
Management, Information Management Strategy, Data
Quality, Metadata Management and Business
Intelligence.
He is a Director of DAMA- I, holds the CDMP Master
certification, is an examiner for CDMP, a Fellow of the
Chartered Institute of Management Consulting (now IC)
a member of the MPO, and SME Director of the DM
Board.
A recognised thought-leader in Information
Management Chris is the author of numerous papers,
books, including sections of DMBoK 2.0, a columnist, a
frequent contributor to industry publications and
member of several IM standards authorities.
He leads an experts channel on the influential
BeyeNETWORK, is a sought after speaker at major
international conferences, and is the co-author of “Data
Modelling For The Business – A Handbook for aligning the
business with IT using high-level data models”. He also
blogs frequently on Information Management (and
motorsport).

Recent Presentations
Enterprise Data World: (DataVersity), May 2014, Austin, Texas, “MDM Architectures &
How to identify the right Subject Area & tooling for your MDM strategy”
E&P Information Management Dubai: (DMBoard),17-19 March 2014, Dubai, UAE
“Master Data Management Fundamentals, Architectures & Identify the starting Data
Subject Areas”
DAMA Australia: (DAMA-A),18-21 November 2013, Melbourne, Australia “DAMA
DMBoK 2.0”, “Information Management Fundamentals” 1 day workshop”
Data Management & Information Quality Europe:
(IRM Conferences), 4-6 November 2013, London, UK
“Data Modelling Fundamentals” ½ day workshop:
“Myths, Fairy Tales & The Single View” Seminar
“Imaginative Innovation - A Look to the Future” DAMA Panel Discussion
IPL / Embarcadero series: June 2013, London, UK, “Implementing Effective Data
Governance”
Riyadh Information Exchange: May 2013, Riyadh, Saudi Arabia,
“Big Data – What’s the big fuss?”
Enterprise Data World: (Wilshire Conferences), May 2013, San Diego, USA, “Data and
Process Blueprinting – A practical approach for rapidly optimising Information Assets”
Data Governance & MDM Europe: (IRM Conferences), April 2013, London, “Selecting
the Optimum Business approach for MDM success…. Case study with Statoil”
E&P Information Management: (SMI Conference), February 2013, London,
“Case Study, Using Data Virtualisation for Real Time BI & Analytics”
E&P Data Governance: (DMBoard / DG Events), January 2013, Marrakech, Morocco,
“Establishing a successful Data Governance program”
Big Data 2: (Whitehall), December 2012, London, “The Pillars of successful knowledge
management”
Financial Information Management Association (FIMA): (WBR), November 2012,
London; “Data Strategy as a Business Enabler”
Data Modeling Zone: (Technics), November 2012, Baltimore USA
“Data Modelling for the business”
Data Management & Information Quality Europe: (IRM), November 2012, London; “All
you need to know to prepare for DAMA CDMP professional certification”
ECIM Exploration & Production: September 2012, Haugesund, Norway:
“Enhancing communication through the use of industry standard models; case study in
E&P using WITSML”
Preparing the Business for MDM success: Threadneedles Executive breakfast briefing
series, July 2012, London
Big Data – What’s the big fuss?: (Whitehall), Big Data & Analytics, June 2012, London,
Enterprise Data World International: (DAMA / Wilshire), May 2012, Atlanta GA,
“A Model Driven Data Governance Framework For MDM - Statoil Case Study”
“When Two Worlds Collide – Data and Process Architecture Synergies” (rated best workshop
in conference); “Petrochemical Information Management utilising PPDM in an Enterprise
Information Architecture”
Data Governance & MDM Europe: (DAMA / IRM), April 2012, London,
“A Model Driven Data Governance Framework For MDM - Statoil Case Study”
AAPG Exploration & Production Data Management: April 2012, Dead Sea Jordan; “A Process
For Introducing Data Governance into Large Enterprises”
PWC & Iron Mountain Corporate Information Management: March 2012, Madrid;
“Information Management & Regulatory Compliance”
DAMA Scandinavia: March 2012, Stockholm,
“Reducing Complexity in Information Management” (rated best presentation in conference)
Ovum IT Governance & Planning: March 2012, London;
“Data Governance – An Essential Part of IT Governance”
American Express Global Technology Conference: November 2011, UK,
“All An Enterprise Architect Needs To Know About Information Management”
FIMA Europe (Financial Information Management):, November 2011, London; “Confronting
The Complexities Of Financial Regulation With A Customer Centric Approach; Applying IPL’s
Master Data Management And Data Governance Process In Clydesdale Bank “
Data Management & Information Quality Europe: (DAMA / IRM), November 2011, London,
“Assessing & Improving Information Management Effectiveness – Cambridge University Press
Case Study”; “Too Good To Be True? – The Truth About Open Source BI”
ECIM Exploration & Production: September 12th 14th 2011, Haugesund, Norway: “The Role
Of Data Virtualisation In Your EIM Strategy”
Enterprise Data World International: (DAMA / Wilshire), April 2011, Chicago IL; “How Do You
Want Yours Served? – The Role Of Data Virtualisation And Open Source BI”
Data Governance & MDM Europe: (DAMA / IRM), March 2011, London,
“Clinical Information Data Governance”
Data Management & Information Management Europe: (DAMA / IRM), November 2010,
London,
“How Do You Get A Business Person To Read A Data Model?
DAMA Scandinavia: October 26th-27th 2010, Stockholm,
“Incorporating ERP Systems Into Your Overall Models & Information Architecture” (rated best
presentation in conference)
BPM Europe: (IRM), September 27th – 29th 2010, London,
“Learning to Love BPMN 2.0”
IPL / Composite Information Management in Pharmaceuticals: September 15th 2010, London,
“Clinical Information Management – Are We The Cobblers Children?”
ECIM Exploration & Production: September 13th 15th 2010, Haugesund, Norway: “Information
Challenges and Solutions” (rated best presentation in conference)
Enterprise Architecture Europe: (IRM), June 16th – 18th 2010, London: ½ day workshop; “The
Evolution of Enterprise Data Modelling”

Recent Publications
Book: “Data Modelling For The Business – A Handbook for aligning the business with IT using high-level data models”; Technics Publishing;
ISBN 978-0-9771400-7-7; http://www.amazon.com/Data-Modeling-Business-Handbook-High-Level
White Paper: “Information is at the heart of ALL Architecture disciplines”,; March 2014
Article: The Bookbinder, the Librarian & a Data Governance story ; July 2013
Article: Data Governance is about Hearts and Minds, not Technology January 2013
White Paper: “The fundamentals of Information Management”, January 2013
White Paper: “Knowledge Management – From justification to delivery”, December 2012
Article: “Chief INFORMATION Officer? Not really” Article, November 2012
White Paper: “Running a successful Knowledge Management Practice” November 2012
White Paper: “Big Data Projects are not one man shows” June 2012
Article: “IPL & Statoil’s innovative approach to Master Data Management in Statoil”, Oil IT Journal, May 2012
White Paper: “Data Modelling is NOT just for DBMS’s” April 2012
Article: “Data Governance in the Financial Services Sector” FSTech Magazine, April 2012
Article: “Data Governance, an essential component of IT Governance" March 2012
Article: “Leveraging a Model Driven approach to Master Data Management in Statoil”, Oil IT Journal, February 2012
Article: “How Data Virtualization Helps Data Integration Strategies” BeyeNETWORK (December 2011)
Article: “Approaches & Selection Criteria For organizations approaching data integration programmes” TechTarget (November 2011)
Article: Big Data – Same Problems? BeyeNETWORK and TechTarget. (July 2011)
Article “10 easy steps to evaluate Data Modelling tools” Information Management, (March 2010)
Article “How Do You Want Your Data Served?” Conspectus Magazine (February 2010)
Article “How do you want yours served (data that is)” (BeyeNETWORK January 2010)
Article “Seven deadly sins of data modelling” (BeyeNETWORK October 2009)
Article “Data Modelling is NOT just for DBMS’s” Part 1 BeyeNETWORK July 2009 and Part 2 BeyeNETWORK August 2009
Web Channel: BeyeNETWORK “Chris Bradley Expert Channel” Information Asset Management
http://www.b-eye-network.co.uk/channels/1554/
Article: “Preventing a Data Disaster” February 2009, Database Marketing Magazine

Information Management Training &
MentoringWe offer a number of training options & Custom-built, on-site training & awareness seminars
can also be delivered.
The following training courses are available:
• Information Management Fundamentals – 5 day introductory course covering all of the components of
Information Management as defined in the DAMA Body of Knowledge (DMBoK) & forthcoming changes in DMBoK 2.0
presented by one of the DMBoK 2 authors
• Data Modelling fundamentals – 3 day intermediate course introducing students to data modelling, its purpose, the different types of models and how to
construct and read a data model.
• Advanced Data Modeling – 3 day advanced course for students with data modelling experience to understand the advanced concepts and human centric
aspects of data modelling to enable them to build quality models that meet business needs.
• IM Fundamentals & Practioner Courses– A series of 1 day (foundation) and 2 day (practitioner) classes to give practitioners a solid background in a specific
Information Management topics. The 2 day practitioner workshops explore more detail on the implementation aspects of the particular Information
Management discipline
• Data Modelling Foundation (1 day only)
• Data Governance Foundation & Practitioner
• Master & Reference Data Foundation & Practitioner
• Data Quality Management Foundation & Practitioner
• Data Warehouse & Business Intelligence Foundation & Practitioner
• Data Integration Foundation & Practitioner
• Executive Workshops – ½ and 1 day executive workshop(s) designed to give non-technical managers a basic understanding of a various Information
Management topics and their importance to the organisation.
• CDMP Certification– 3 day workshop “exam cram” designed to help attendees pass the DAMA CDMP certification. Sitting the live examinations is included
as part of the workshop.
• Integrated Business Process, Data & Requirements Definition– 5 day intensive class to show students an integrated requirements discovery and
definition approach covering business process, different types of requirements modelling, and the critical role of the conceptual data model.

Data Modelling 101
› CONTEXT WITHIN THE DMBOK
› DATA & METADATA
› DATA MODELLING: WHAT & WHY?
› TYPES & LEVELS OF DATA MODELS
› DATA MODEL COMPONENTS
› NORMALISATION
› DIMENSIONAL DATA MODELLING
› IT’S NOT JUST FOR DBMS’S
› SUMMARY

DATA
ARCHITECTURE
MANAGEMENT
DATA
DEVELOPMENT
DATABASE
OPERATIONS
MANAGEMENT
DATA SECURITY
MANAGEMENT
REFERENCE &
MASTER DATA
MANAGEMENT
DATA QUALITY
MANAGEMENT
META DATA
MANAGEMENT
DOCUMENT & CONTENT
MANAGEMENT
DATA
WAREHOUSE
& BUSINESS
INTELLIGENCE
MANAGEMENT
DATA
GOVERNANCE
› Enterprise Data Modelling
› Value Chain Analysis
› Related Data Architecture
› External Codes
› Internal Codes
› Customer Data
› Product Data
› Dimension Management
› Acquisition
› Recovery
› Tuning
› Retention
› Purging
› Standards
› Classifications
› Administration
› Authentication
› Auditing
› Analysis
› Data modelling
› Database Design
› Implementation
› Strategy
› Organisation & Roles
› Policies & Standards
› Issues
› Valuation
› Architecture
› Implementation
› Training & Support
› Monitoring & Tuning
› Acquisition & Storage
› Backup & Recovery
› Content
Management
› Retrieval
› Retention
› Architecture
› Integration
› Control
› Delivery
› Specification
› Analysis
› Measurement
› Improvement
Where does
Data Modeling
fit?

Is there more to life
than this?
is ais ais ais ais ais ais ais ais a
owns
is assigned
is owner of is owned by
belongs to has
registered at
classifies
has
classifies
transacted in
parent of
assigns assigns
performs
determines
payee-correspondent relationship
is granted
has
has
requires
Business Party
Business Person
Business Partner
Legal Entity
Address Type
Address
Bank Account
Contact
Counterparty Transportation Provider Financial Institution Exchange Inspector Broker Clearing HouseRegulatory Agency Rating Agency
Currency
Industry Classification
Industry Classification Scheme
Legal Entity Identification
Legal Entity Identifying Scheme
Bank Account Type
Internal Operating Unit
LE Ownership
Indirect Tax Registration
Tax Exemption License
Contact Usage
Contact Role
Payment Security Type

What is Data?
FIGURES FACTS
THINGS
OBSERVED
FACTS
Current Balance = $400
Home Town = Bath
Order Placed Date = 4th November 2013
Customer id = 987654321
Person image = Data is objective
Data in context = Information

Data
Person id Forename Family Name Salutation NI
number
Player
Rating
Office address Credit
limit
Emergency
Contact
12356 Mitchell Stark Mr 112113 2 123 St James Place,
London, WC1
$0.25m F.J Banks
124 Gary Sobers Sir 112141 2 Shellmex house, London,
EC1
$0.25m E.C. Dollar
09211 Alan Knott Mr 201221 4 IBM House, White Plains,
NY
$0.35m F.E. Goodwin
43219 Rachel Hahoe-Flint Mrs 202119 5 Microsoft, MS Business
Park, Seattle, WA
$0.55m R.B. Gibb
12 Allan Border Mr 311456 5 Dell park, Palo Alto, CA $0.5m S.T.Law
230 Ian Botham Sir 429876 0 Seattle Aero Park,
Seattle, WA
- -

What is Metadata?
Metadata
is .... data
It is,
however,
data
ABOUT
data
Person id: This is the unique identification number
for the customer as used in our organisation.
Forename: The preferred name used by the
person. Note this is not the same as the birth
certificate forename.
NI Number: The National Insurance number for the
person.
THIS is the
difference
Metadata is
context
dependent
KEY point

Metadata
Description = This
is the unique
identification
number for the
person as used
in our
organisation.
Size = 1-5
numeric
Mandatory =
Yes
Unique key =
Yes
PERSON ID
Metadata has
“properties”
These describe the
characteristics & rules
of the metadata

Where do you encounter
metadata every day?

MetaData
DATA METADATA

Where else do you use
metadata every day?

Exercise 1:
Data or Metadata
DATA OR METADATA?
Chris Bradley
Company Name
750 metres
1 Royal Crescent, Bath
Shell
Dubai
Date Of Birth
Order Date
Nov 3rd 2014
Order Status
Chris.bradley@DMAdvisors.co.uk
+44 7808 038 173
Location
Singer Name

Exercise 1:
Data or Metadata
DATA OR METADATA?
Chris Bradley
Company Name
750 metres
1 Royal Crescent, Bath
Shell
Dubai
Date Of Birth
Order Date
Nov 3rd 2014
Order Status
Chris.bradley@DMAdvisors.co.uk
+44 7808 038 173
Location
Singer Name
Data
Data
Data
Data
Data
Data
Data
Data
Data
Data
Metadata
Metadata
Metadata
Metadata
Metadata
Metadata

What Is Data Modelling?

What Is A Data Model?
A model is a
representation of
something in our
environment making
use of standard
symbols to enable
improved
understanding of the
concept
A data model
describes the
specification,
definition and rules
for data in a business
area
A data model is a
diagram (with
additional supporting
metadata) that uses
text and symbols to
represent data to
give the reader a
better understanding
of the data
A data model
describes the
inherent logical
structure of the data
within a given
domain and, by
implication, the
underlying structure
of that domain itself

What Is A Conceptual Data Model?
 A description of a Business (or
an area of the Business) in
terms of the things it needs to
know about.
 The Data things are “entities”
and the “facts about things”
are attributes & relationships.
 It’s a representation of the
“real world”, not a technical
implementation of it
 Should be able to be
understood by Business users
Definition:
A Student is any person who has been admitted to a course, has paid, and has enrolled in one or more
modules within a course. Tutors and other staff members may also be Students
Business Assertions
 A Student enrolls for one or more modules
 A Course can be taught through one or more
Modules
 A Room can be the location of one or more
modules
 A Tutor can be the teacher of one or more
modules
The Other Way?
 A Module is enrolled in by many students
 A Module is an offering within one course
 A Module is located in one room
 A Module is taught by one tutor
Really?

A Data Model Represents
Person, Employee, Vendor, Customer,
Department, Organisation, …WHO
Product, Service, Raw Material, Training
Course, Flight, Room, …WHAT
Time, Day, Date, Calendar, Reporting Period,
Fiscal Period, …WHEN
Geographic location, Delivery address,
Storage Depot, Airport, …WHERE
Order, Complaint, Inquiry, Transaction, …WHY
Invoice, Policy, Contract, Agreement,
Document, Account, …HOW
Classes of
entities
(kinds of things)
about which a
company
wishes to know
or hold
information

What is an Entity?
Entity: A classification of the types of objects found in the real world --
persons, places, things, concepts and events – of interest to the
enterprise. 1
1 DAMA Dictionary of Data Management
WHO? WHERE?
WHEN? HOW?WHY?
WHAT?
The “Who, What, Where, When, Why” of the Organization

Identifying Entities
Is it an
Entity?
Does this imply an instance of a SINGLE
thing, not a group or collection
How do I identify ONE of those things?
What are the facts I want to hold against
ONE of those things?
Do I even WANT to hold facts about
these things?
PROCESSES will act upon it, so does the
“thing” make sense in a well formed
process phrase i.e. a verb – noun pair?
What is ONE of those things?

Sample Entities
Product
Customer
Location
Order
Raw
Material
Ingredient
Region

Exercise 2: Entities
Student Building Maths Department
Course
Catalogue
Attendance
Sheet
Enrolment
Form
Professor
Plumb
Prerequisite
list
Module
Organisation
Chart
Student
Directory
Module
Description
Qualification
Certification
Body
Graduation
Which of these might / might not be valid entities?

PERSON ID
FIRST NAME
DATE OF BIRTH
PRODUCT NUMBER
QUANTITY ORDERED
FLIGHT NUMBER
SEAT CLASS
…
the
attributes
of that
information
(facts about
things)

Attributes An Attribute is a piece of information
about or a characteristic of an Entity.
Attributes
Entity
Employee • Employee Identifier
• Employee Last Name
• Employee First Name
• Employee Hire Date
• Employee Signed Employment Contract
• Employee Drivers License Photo
Entity
Attributes

Attributes
 Facts about “entities” are
recorded as attributes &
relationships.
 We don’t record every fact,
only the ones that are needed
Attribute Properties
 “user-entered” vs. “constrained set”: The
attribute can only come from a finite set,
such as code list / drop down set
 fundamental vs. redundant: the same
value is recorded multiple times in different
entities
 single-valued vs. multivalued: one
attribute can have multiple values, at a
time or over time
Ok for a
conceptual
data
model

“Each CUSTOMER
is the placer of
zero, one or more
ORDER(s)"
Relationships should be
named in both directions,
thus in the other direction
we have:
"Each ORDER
must be placed
by one and only
one CUSTOMER"
A relationship called "is the
placer of" operates on entity
classes CUSTOMER and
ORDER and forms the
following concrete assertion:
Is this true…
always?
Is this true?
relationships
among those
entities and
(often implicit)
relationships
among those
attributes
Relationships
form a concrete
Business Assertion

A Data Model represents
It’s much more than a
picture!
Classes of
entities
(kinds of things)
about which a
company
wishes to know
or hold
information
the
attributes
of that
information
(facts about
things)
relationships
among those
entities and
(often implicit)
relationships
among those
attributes
The model
describes the
organization of
the data
irrespective of
how data might
be represented in
a computer
system

Why Produce A Data Model?
TOP REASONS*
1. Capturing Business Requirements
2. Promotes Reuse, Consistency, Quality
3. Bridge Between Business and Technology
Personnel
4. Assessing Fit of Package Solutions
5. Identify and Manage Redundant Data
6. Sets Context for Project within the
Enterprise
7. Interaction Analysis: Complements
Process Model
8. Pictures Communicate Better than Words
9. Avoid Late Discovery of Missed
Requirements
10. Critical in Managing Integration Between
Systems
11. Pre-cursor to DBMS design / generate
DDL
* DAMA-I Survey

Why Data Modelling Is Important
Provides a
common
vocabulary
An aid to
understanding
Don’t need to
look at the detail
right away (or
sometimes ever)
Needs to be
understood to be
managed
Data is an asset
of your
corporation
Context and high
level views will be
sufficient
Provide only the
level of detail
that is necessary
– fit for purpose

Why Data Modelling Is Important
BUSINESS
ARCHITECTURE
Business
Objectives & Goals
Motivations &
Metrics
Functions, Roles,
Departments
INFORMATION
ARCHITECTURE
Enterprise Data
Model
Conceptual Data
Models
Logical Data
Models
Physical Data
Models
PROCESS
ARCHITECTURE
Overall Value
Chain
High-Level
Business Processes
Workflow Models
APPLICATION / SYSTEMS
ARCHITECTURE
Systems within
Scope
High-Level Mapping
Business Services
Presentation
Services (use cases)

Why Data Modelling Is Critical
BUSINESS
ARCHITECTURE
Business Objectives &
Goals
Motivations & Metrics
Functions, Roles,
Departments
INFORMATION
ARCHITECTURE
Enterprise Data Model
Conceptual Data
Models
Logical Data Models
Physical Data Models
PROCESS
ARCHITECTURE
Overall Value Chain
High-Level Business
Processes
Workflow Models
APPLICATION / SYSTEMS
ARCHITECTURE
Systems within Scope
High-Level Mapping
Business Services
Presentation Services
(use cases)
The company is undertaking
a radical approach to
enhance Customer
experience, service and
satisfaction by providing
seamless multi-channel
Customer access to all core
services
B U S I N E S S
O B J E C T I V E S
I N F O R M A T I O N
S E R V I C E S
B U S I N E S S
S E R V I C E S
PRESENTATION SERVICES
B U S I N E S S
P R O C E S S
ALL of the Architecture disciplines use the language
(and rules) of the data model

Framework for Enterprise Information Management
EIM goals and strategies are business-driven for
the entire enterprise, underpinned by guiding
principles supported by senior management
Roles, responsibilities, structures and
procedures to ensure that data assets are
under active stewardship
Processes, procedures and
policies to ensure data is fit for
purpose and monitored
Metadata capture, management,
& manipulation to place data in
business & technical context
Proactive planning for the information lifecycle
including the acquisition, manipulation, access,
use, archiving & disposal of information
The identification of appropriate data
integration approach for business challenges
e.g. ETL, P2P, EII, DV, EII or EAI
The full lifecycle control and management
of information from acquisition to
retention & destruction
Organise information to align with business
& technical goals using Enterprise,
Conceptual, Logical & Physical models
Management of Data Warehouses and creation
of actionable Business Intelligence to provide
intelligence and analytics for business benefit
Identification, management and delivery to
consuming applications of the core shared data
concepts required enterprise wide
Manage diverse data sources across the
organisation from transaction data
management, to data warehousing and
business intelligence, to Big Data analytics
Development of realistic Information Management strategies to align the desired Information capabilities and services with business motivations and
strategies. The information initiatives can be accelerated by use of our Reference Architecture models to understand the capabilities, and typical functional
areas for each IM discipline under consideration (such as MDM, DQ, Data Integration etc.). Our Architecture Reference models contain the typical areas of
functionality & capabilities observed in each IM discipline. Our EIM framework has capability & maturity models for each of the IM disciplines together with
the typical processes and activities observed in mature organisational services for each.

Data Model Levels
E N T E R P R I S E
C O N C E P T U A L
L O G I C A L
P H Y S I C A L
S Y S T E M
IMPLEMENTATIONFOCUS
COMMUNICATIONFOCUS

Levels of Data Models
Enterprise
Data
Model
Documents the very
high level business
data objects and
definitions. Enterprise
wide scope to
provide a strategic
view of Enterprise
data.
Conceptual
Data Model
(Subject area)
The business key,
attributes and
definitions of business
data objects. Also
shows the
relationship between
business data
objects. Broader
scope than LDM and
may cover a subject
area (also known as
subject area data
model).
Logical
Data Model
(Application)
Documents the
business key,
attributes and
definitions of business
data objects. It also
shows the
business data
objects. Frequently is
within the scope of a
defined project.
Physical
Data
Model
Technical design eg
tables, columns, keys,
foreign keys, and
other constraints to
be implemented in
the database or
XSD. May be
generated from a
logical data model.
This model is within
the scope of a
defined project.

Enterprise
Conceptual
Logical
Physical
Enterprise vs. Conceptual
vs. Logical
Agree basic
concepts and rules
Detail, may lead
to physical design
Big picture
Optimised for specific
technical environment
DIFFERENTPERSPECTIVESANDLEVELS
OFDETAILFORDIFFERENTUSES
› Common understanding before progressing too far into detail
› Used to communicate with the Business
› Overview: main entities, super types, attributes, and relationships
› Lots of Many to Many & multi meaning relationships
› Relationships frequently show multiplicity of meaning
› May be denormalised
› Non-atomic & multi-valued attributes allowed; no keys
› Should fit on one page
› 20% of the modelling effort
› Detailed: ~ 5x Entities vs Conceptual model
› Detailed: Frequently pre-cursor to 1st cut physical (database) design
› Detailed: Key input to requirements specification
› M:M relationships resolved: Intersection entities mostly have meaning
› Relationship optionality added
› Primary, foreign, alternate keys included
› Reference entities included
› Fully normalized – no multi-valued, redundant, non-atomic attributes
› May be partitioned (sub-models)
› 80% of the modelling effort
CONCEPTUAL/LOGICALKEYDIFFERENCES

Different Data Models For
Different Audiences: Business

Different Data Models For
Different Audiences: Technical

Logical Data
Modelling
Components
Attributes
Entities
Relationships

Logical Data Model Components
Entity Non identifying
relationship
Primary
Key (PK)
Foreign
Key (FK)
AttributesIdentifying
relationship
Recursive
relationship

Person, Employee, Vendor, Customer,
Department, Organisation, …WHO
Product, Service, Raw Material, Training
Course, Flight, Room, …WHAT
Time, Day, Date, Calendar, Reporting Period,
Fiscal Period, …WHEN
Geographic location, Delivery address,
Storage Depot, Airport, …WHERE
Order, Complaint, Inquiry, Transaction, …WHY
Invoice, Policy, Contract, Agreement,
Document, Account, …HOW
Entities
A THING OF SIGNIFICANCE TO THE BUSINESS ABOUT
WHICH INFORMATION NEEDS TO BE KNOWN OR HELD
ABSTRACT
REAL
INDEPENDENT
DEPENDENT

Entity Naming Best Practice
Can be accomplished by placing qualifiers or quantifiers
in front of the entity name.
Entity names must be unique
(e.g. Order)
Use a noun alone wherever possible (e.g. Contract)
Entity names must have
business meaning
e.g. Lease Contract or Back Order
Use adjective + noun or,
adjective + adjective + noun
to clarify meaning
e.g. “Product Group” not “PRODUCT GROUP”
Entity names should be in
Title Case format
The entity is defined in terms of a single occurrence.
(e.g. “Product” not “Products”)
Entity names should be
singular
Do not use “_” within Entity names, this really is only
needed for implementation artefacts (e.g. Tables)
Acronyms or abbreviations
should not be used

Entity Definition Best Practice
› Can the term be understood by
reading the definition?
› Is it unambiguous?
› Avoid jargon & stating the obvious
CLARITY
› Appropriate level of detail – not too
generic – not too specific
› Goldilocks principle
› Contains all necessary components
without omission (e.g. derivations, UOM)
COMPLETENESS
› A subject matter expert would agree
› Relevant to the state of the entity:
i.e. the stages an entity may go
through over time
ACCURACY
Suspect
Prospect
New
Customer
Customer
Silver
Customer
Gold
Customer
Dormant
Customer
Lapsed
Customer

Common Errors
With Entities
› Creating an Entity that is really a report or a
screen or form
› Failing to clarify if the entity deals with types (or
categories) vs. specific instances of things
› Identifying an entity that exists in the real world,
but whose instances can't be uniquely identified
e.g. “Billboard observer”
› Identifying entities that are too imprecise and / or
the name doesn't' imply a single instance
e.g. “Weather”
› Crossing to the “techno side” and introducing
implementation specific constructs
A few common mistakes are regularly
encountered when creating business
focused data models:

Dependent
Why is it important to know
whether an entity is independent
or dependent?
You don’t need to know
anything about a Product to
identify a Customer
You don’t need to know
anything about a Customer
to identify a Product
Can be identified without
reference to another entity
on the model
Does not depend on any
other entity for its existence.
3 types:
1. Attributive (depends only
on its parent
2. Associative (depends upon
two or more entities)
3. Category (AKA supertype)
Can depend upon
Independent or Dependent
entities
Depends upon one (or more)
other entities
Independent

PRIMARY KEY (PK)
› Identifies an
entity/table for the
system.
› May be “natural” or
“surrogate”
› Could be composite
ALTERNATE KEY (AK)
› Another way to identify
an entity/table.
› May be surrogate
FOREIGN KEY (FK)
› Identifies a relation
among entities/tables
INVERTED KEY (IK) OR
INVERSION ENTITY [IE]
› Improves access to
table information
(physical)
Attributes & Keys
KEYS (SUMMARY):
AN ENTITY CAN CONTAIN FOUR TYPES OF KEYS
EntityKeys
Composite is made up of several items to form the key
Surrogate is often a
“made up” key

May Indicate Relationship Membership (FK)
Attribute (Business Not Designer Added)
Unchanging
Must Uniquely Identify Each Instance Of
The Entity
Mandatory
Can I use
Registration
Number as the
PK?
Primary Key
Attribute(s) that make up a PK are
represented in modelling tools separately
from the rest of the attributes by a line

Primary Keys
› What attributes might uniquely identify an entity? Let’s use Customer as an example.
› What might uniquely identify an individual customer?
Is Last Name + First Name enough?
— Could there be 2 customers named John Smith?  Probably
Is Last Name + First Name + Date of Birth enough?
— Could there be 2 customers named John Smith born on 1 June, 1963?  Less
Likely, but Possible
Is Last Name + First Name + Date of Birth + Address enough?
— Could there be 2 customers named John Smith born on 1 June, 1963 living at 1
Earl’s Court, London, UK?  Even Less Likely. Possible, but how many attributes
do we want to use?

Keys:
Natural vs. Surrogate
› The “Customer” example keys we just
identified would be classified as natural
keys.
› Natural keys are based on business rules
and logic that determine how an individual
instance can be uniquely identified.
› As we’ve seen, natural keys can become
unwieldy, requiring a number of attributes,
which makes queries difficult.
› Also, extreme care is needed as
components of natural keys could change
— Surrogate keys are often used instead,
which are system-generated unique
identifiers. e.g. Customer ID, Product ID, etc.
— While surrogate keys are more efficient,
important business rules are lost when they
are used. It’s a balancing act.

Alternate Key
ALTERNATE KEY (AK):
ANOTHER WAY OF IDENTIFYING THE ENTITY
Are these good
AK’s for all
circumstances?

Foreign Key
FOREIGN KEY (FK):
POINTS BACK TO THE PARENT(S)
Can a FK be
part of the
Primary Key?

Inverted Key INVERTED KEY (IK):
PHYSICAL IMPLEMENTATION ONLY
Primary
Key
Car Eye
Colour
Nationality Salary
1 Audi Blue UK 100000
2 BMW Green US 95000
3 BMW Brown FR 85000
4 Bentley Blue UK 250000
5 Audi Brown US 60000
6 Ford Blue FR 50000
7 Ford Brown UK 45000
8 Audi Brown US 55000
9 BMW Blue UK 65000
TABLE
Audi (1, 5, 8)
Bentley (4)
BMW (2, 3, 9)
Ford (6, 7)
Blue (1, 4, 6, 9)
Brown (3, 5, 7, 8)
Green (2)
FR (3, 6)
UK (1, 4, 7, 9)
US (2, 5, 8)
CAR IK EYE COLOUR IK NATIONALITY IK
1 2 3 4 5 6 7 8 9
Car = Audi 1 1 1
Car = Bentley 1
Car = BMW 1 1 1
Car = Ford 1 1
Eyes = Blue 1 1 1 1
Eyes = Brown 1 1 1 1
Eyes = Green 1
Nationality = FR 1 1
Nationality = UK 1 1 1 1
Nationality = US 1 1 1
IK IMPLEMENTED AS BITMAP INDEX

E.g.
Person Id
“The unique identifier of a Person in
Acme corporation. This is a unique
integer length 13, with digit 13
being a checksum”
Employee Id:
Domain name = Person Id
Business Sponsor Id:
Cost Centre Owner:
Customer Id:
Customer Name:
Domain name = Person Name
DOMAINS CAN BE CASCADED
TO OTHER ATTRIBUTES
Characteristics of
an attribute
THE COMPLETE SET OF VALID
VALUES A DATA ELEMENT MAY
CONTAIN (EG DROP DOWN)
Attribute Properties & Domains
Name
• Logical
• Physical
Unique
Part of primary key
Mandatory
Datatype
Domain
Validation Rules
Default Value
Nullable (Y/N)
Definition
Notes
MainAttributeProperties Why are
Domains
useful?

Relationships
Between Entities
When resolving Many to
Many relationships,
the “Intersection Entity”
invariably has real
business meaning
Well formed
relationships
represent a business
assertion that can be
tested:
“Is it true that an Order
can only be placed by
one Customer?”
A relationship
represents an association
between two entities,
ensuring the referential
integrity among instances
of the entities.

What does this
tell us about
Asset Id?
Relationship Types
TYPES OF “ONE TO MANY” RELATIONSHIPS:
MANDATORY, NON-IDENTIFYING
Denotes mandatory
(by the entity that is mandatory)
Denotes non-
identifying
Parent id IS NOT part of
the child unique id. Just a
foreign key.

Relationship Types
2 parts in this PK
Installation has 2 identifying
mandatory relationships to it
1 part in this PK
Asset has 1 non-identifying
relationship to it
Think of
another
example

This tells us that the Order
Number IS unique
Relationship Types

What does this
tell us about
Asset Id?
Relationship Types
MANDATORY, IDENTIFYING
Denotes mandatory
(by the entity that is mandatory)
Denotes
identifying
Parent id becomes a part
of the child unique id
Think of
another
example

Installation has 2
identifying mandatory
relationships to it
Relationship Types
MANDATORY, IDENTIFYING
Now there’s 2 parts in this PK
Asset has 1 identifying
mandatory relationship to it

What does this
tell us about
Asset Id?
Relationship Types
OPTIONAL, NON-IDENTIFYING
Denotes optional
(by the entity that is optional)
Denotes non-
identifying
Parent id IS NOT part of
child unique id. Just a
foreign key.

Relationship Types
Installation still has 2 identifying
mandatory relationships to it but the
primary key has changed as it now
doesn’t include Equipment Type Code
Think of
another
example

This tells us that we don’t
need to know the
Organisation to find a
Person. A Person does NOT
have to be “employed by”
an Organisation.
Relationship Types
Back to
our Asset
example
Organisation to Person
relationship is optional,
non-identifying

So what
have we
got to do
now?
Relationship Types
See what happens to
Installation if its 2
relationships become
non-identifying.

Why can’t a many
to many relationship
be an “identifying”
relationship?
Chris’s law
99% of M:M relationships
represent a real business
concept that is the
intersection entity
Relationship Types
MANY TO MANY (AKA NON SPECIFIC)
Think of
another
example

What’s the real
business “thing”
that resolves this
many to many?
MANY TO MANY (AKA NON SPECIFIC)
Qualification to
Person relationship is
non-specific (AKA
many to many)
Exercise 3: Relationships
Think of 2 more
common M:M
examples

What’s the real
business “thing”
that resolves this
many to many?

Recursive Relationships
Can a recursive
relationship be
“identifying”?
It is possible to have
many recursive
relationships between
the same entity and
itself.
But we cannot
have duplicated
Attribute Names in an
Entity hence the FK
must have a role (e.g.
Managed by).
A recursive
relationship occurs
when there is a
an entity and itself.
Can a recursive
relationship be
“mandatory”?

Entity Subtypes Super-type Entity:
Contains the primary
key and common
attributes
Discriminator: Attribute
to determine which
subtype we are talking
about
Sub-type Entities:
Contain Specific
attributes for each type

Normalisation
WHY NORMALISE A DATA MODEL?
Improved
understanding
Ensure data
integrity
Easier to
query
Remove
data
redundancy
Improved data quality
Reduction in
timescales
Easier
maintenance
LEADING
TO

Normalisation Approaches
A data model is fully
normalised when it is in Third
Normal Form (3NF).
• 3NF is a normalisation method
indicating 3 stages of
normalisation:
Further normalisation
methods can also be
applied for very specific
cases. See advanced
course for details:
• Boyce/Codd normal form (BCNF)
• Fourth Normal Form (4NF)
• Fifth Normal Form (5NF)

1st
Normal Form
A P R I M A R Y K E Y M U S T B E
› Unique - the primary key uniquely identifies
each instance of the entity
› Mandatory – the primary key must be defined
for every instance of the entity
› Unchanging – while not mandatory, it is
desirable that the primary key does not change
T O P U T A M O D E L I N T O 1 N F
› Identify the primary key
› Remodel repeating values
› Remodel multi-valued attributes
1NF DEFINITION:
EVERY NON-KEY ATTRIBUTE IN AN ENTITY MUST
DEPEND ON IT’S PRIMARY KEY

1st Normal
Form Example
NAME GENDER EMAIL ADDRESS
Barack Obama Male barack@whitehouse.org
gobama@vote2012.com
judgementday@dontnuke.com
David Cameron Male Callmedave@tory.co.uk
DC@uk.gov.com
Angela Merkel Female FrauAng@gov.de
Julia Gillard Female madeinwales@gov.au
julia@outofwork.com
To put this in
first normal
form we must:
Identify a
primary key
Remodel the
multi-valued
attribute of Name
Remodel the multi-
valued attribute of
Email Address

A P R I M A R Y K E Y M U S T B E
› Unique - the primary key uniquely identifies
each instance of the entity
› Mandatory – the primary key must be defined
for every instance of the entity
› Unchanging – while not mandatory, it is
desirable that the primary key does not change
1st Normal
Form Example
What is the
primary key:
Name,
Gender, Email?
So there is no primary key!
So what do we do…?
We create our own
primary key (virtual key)
To put this in
first normal
form we must:
Identify a
primary key
Remodel the
multi-valued
attribute of Name
Remodel the multi-
valued attribute of
Email Address

1st Normal
Form Example
Name = First Name + Last Name
To put this in
first normal
form we must:
Identify a
primary key
Remodel the
multi-valued
attribute of Name
Remodel the multi-
valued attribute of
Email Address
gobama@vote2012.com
DC@uk.gov.com
julia@outofwork.com

To put this in
first normal
form we must:
Identify a
primary key
Remodel the
multi-valued
attribute of Name
Remodel the multi-
valued attribute of
Email Address
1st Normal
Form Example
gobama@vote2012.com
DC@uk.gov.com
julia@outofwork.com
Next we look at this

1NF: Email Address?
› A multi-valued attribute: name & domain?
› Can we identify ALL available types?
» Home Email
» Work Email
» Club Email
› An exhaustive list?
› No; a Person can have any number of email addresses.
› We need to allow for a Person having any number of
email addresses.
Is it just a
compound
attribute?
No, it is
multi-
valued!
To put this in
first normal
form we must:
Identify a
primary key
Remodel the
multi-valued
attribute of Name
Remodel the multi-
valued attribute of
Email Address

1st Normal
Form Example
1NF DEFINITION:
P E R S O N P E R S O N E M A I L
& as it’s a simple example
they are in 2NF & 3NF too!
PERSON
ID
FIRST NAME LAST NAME GENDER
1 Barack Obama Male
2 David Cameron Male
3 Angela Merkel Female
4 Julia Gillard Female
PERSON
ID
EMAIL ADDRESS
1 barack@whitehouse.org
1 gobama@vote2012.com
1 judgementday@dontnuke.com
2 Callmedave@tory.co.uk
2 DC@uk.gov.com
3 FrauAng@gov.de
4 madeinwales@gov.au
4 julia@outofwork.com
We have now put our politicians in 1NF!

1st Normal
Form Example
REGISTRATION MODEL CHASSIS NUMBER MILEAGE FEATURES
HV62SYG Lexus 450H 76365296745568432 7,129 Electric Windows
Satellite Navigation
Bluetooth integration
Head Up Display
Speech Control
Y612 SYG Audi A4 13847621837653275 10,732 Electric Windows
WN09 UTS BMW 320d 32178468273647327 31,123 Electric Windows
Satellite Navigation
WU52XUX Ford Focus 71283459735474924 104,123 Electric Windows
Turn this
into 1NF:
State your
Keys
1NF DEFINITION:

1st Normal
Form Example
CHASSIS NUMBER FEATURE
76365296745568432 Electric Windows
76365296745568432 Satellite Navigation
76365296745568432 Bluetooth integration
76365296745568432 Head Up Display
76365296745568432 Speech Control
C A R E N T I T Y C A R F E A T U R E E N T I T Y
CHASSIS NUMBER REGISTRA-
TION
MANUFACT
-URER
MODEL MILEAGE
76365296745568432 HV62SYG LEXUS RX450H 7,129
13847621837653275 Y612 SYG AUDI A4 10,732
32178468273647327 WN09 UTS BMW 320D 31,123
71283459735474924 WU52XUX FORD FOCUS 104,123
The primary
keys are in
BLUE
But, refer back
to primary key
criteria
Feature names
are likely to
change

1st Normal
Form Example
ID
FEATURE
76365296745568432 1 Electric Windows
76365296745568432 2 Satellite Navigation
76365296745568432 3 Bluetooth integration
76365296745568432 4 Head Up Display
76365296745568432 5 Speech Control
TION
MANUFACT
-URER
MODEL MILEAGE
76365296745568432 HV62SYG LEXUS RX450H 7,129
13847621837653275 Y612 SYG AUDI A4 10,732
32178468273647327 WN09 UTS BMW 320D 31,123
71283459735474924 WU52XUX FORD FOCUS 104,123

2nd
Normal Form
2NF DEFINITION:
EACH ENTITY MUST HAVE THE FEWEST POSSIBLE
CORRECT PRIMARY KEY ATTRIBUTES
For each non-
key attribute
(i.e. not a primary,
foreign or alternate
key)
Test if it
depends
entirely on the
primary key
If it doesn’t,
move it out to a
new entity

2nd Normal
Form Example
Does Feature
depend entirely
upon Chassis
Number and
Feature Id?
ID
FEATURE
76365296745568432 4 Head Up Display
76365296745568432 5 Speech Control
C A R F E A T U R E E N T I T Y

2nd Normal
Form Example
ID
76365296745568432 1
76365296745568432 2
76365296745568432 3
76365296745568432 4
76365296745568432 5
13847621837653275 1
13847621837653275 3
32178468273647327 1
32178468273647327 2
32178468273647327 3
71283459735474924 1
FEATURE
ID
FEATURE
1 Electric Windows
4 Head Up Display
5 Speech Control
TION
MANUFACT
-URER
MODEL MILEAGE
76365296745568432 HV62SYG LEXUS RX450H 7,129
13847621837653275 Y612 SYG AUDI A4 10,732
32178468273647327 WN09 UTS BMW 320D 31,123
71283459735474924 WU52XUX FORD FOCUS 104,123
Car Feature is an
associative entity
We have now put our cars into 2NF
F E A T U R E E N T I T Y

3rd
Normal Form
3NF DEFINITION:
EACH NON KEY ELEMENT MUST BE DIRECTLY
DEPENDENT UPON THE PRIMARY KEY AND NOT
UPON ANY OTHER NON-KEY ATTRIBUTES
For each non-
key attribute
(i.e. not a primary,
foreign or alternate
key)
Test if it
depends
entirely on the
primary key &
nothing else
If it doesn’t,
move it out to a
new entity

3rd Normal
Form Example
TION
MANUFACT
-URER
MODEL MILEAGE
76365296745568432 HV62SYG LEXUS RX450H 7,129
13847621837653275 Y612 SYG AUDI A4 10,732
32178468273647327 WN09 UTS BMW 320D 31,123
71283459735474924 WU52XUX FORD FOCUS 104,123
For 3NF, all attributes must
depend only on Chassis Number.
But “Model” also depends
upon Manufacturer

3rd Normal
Form Example
TION
MODEL
ID
MILEAGE
76365296745568432 HV62SYG 1 7,129
13847621837653275 Y612 SYG 2 10,732
32178468273647327 WN09 UTS 3 31,123
71283459735474924 WU52XUX 4 104,123
MODEL
ID
MANUFACT
-URER ID
MODEL
1 1 RX450H
2 2 A4
3 3 320D
4 4 FOCUS
Its possible that 2 manufacturers may make a
car with the same name. The manufacturer
and model together make a key.
The model name on its own is not a key
candidate since it may not be unique:

3rd Normal
Form Example
TION
MODEL
ID
MILEAGE
76365296745568432 HV62SYG 1 7,129
13847621837653275 Y612 SYG 2 10,732
32178468273647327 WN09 UTS 3 31,123
71283459735474924 WU52XUX 4 104,123
MODEL
ID
MANUFACT
-URER ID
MODEL
1 1 RX450H
2 2 A4
3 3 320D
4 4 FOCUS
MANUFACT-
URER ID
MANUFACT-
URER NAME
CONTACT EMAIL
1 Lexus hitori@lexus.jp
2 Audi hans@audi.de
3 BMW woflgang@bmw.de
4 Ford dwane@ford.com
We have now put our cars into 3NF
3NF DEFINITION:
EACH NON KEY ELEMENT MUST BE DIRECTLY
DEPENDENT UPON THE PRIMARY KEY AND NOT
UPON ANY OTHER NON-KEY ATTRIBUTES

Normalisation Summary
0NF
1NF
2NF
3NF
4& 5NF
Un-normalised (UNF or 0NF)
Contains a “repeating group”
First Normal Form (1NF)
Repeating attributes moved down to associative
entities
Second Normal Form (2NF)
Only applies to dependent entities
No attributes in a child entity are really facts about a
parent (or grandparent). No characteristic or
associative entity redundantly contains facts from its
parent(s) – if it does, move the fact(s) up and if
necessary create a new parent entity
Third Normal Form (3NF)
If any entity redundantly contains facts from a
related (non-parent) entity, move the fact(s) out to
the other entity and create a new entity if necessary
Fourth and Fifth Normal Form (4NF, 5NF)
“Large” (3-way or more) associatives need to be
broken down into more granular entities

› Using visual cues consistently
› Having a starting point and direction
› Abstracting
› Masking unnecessary detail
› Highlighting what matters
“Let’s start here with Special Tax Rate
Variation Comment Type…”
Graphical Principles
OUR MODELS SHOULD AID UNDERSTANDING BY:

Dimensional Data Models
Designed for the rapid recovery
of information to be delivered to
OLAP systems or ad hoc analysis
Modeled based on the
dimensional modeling principles
popularised by Ralph Kimball
“Entity-relationship modelling is a
logical design technique that seeks to
eliminate data redundancy”
A good choice for
OLTP systems
“Dimensional modelling is a design
technique that seeks to present data
in a way that maximises both ease
of use and query performance.”
A good choice for Data
Warehouses / Business
Intelligence systems

Model Features
RELA TIONAL
› Optimised for OLTP
› Normalised
› Low redundancy
› Relationships between entities are
explicit
› Tightly coupled to business model
DIMEN SIONAL
› Optimised for reporting
› Business Entities are denormalised
› More redundancy to support faster
query performance
› Relationships between entities are
implicit
› Loosely coupled to business model

A Dimensional Model
‘STAR SCHEMA’
composed of Dimension and Fact tables
Dimension tables
› EG: Location, Product,
Time, Promotion,
Organisation …
› Product dimension
includes Product Type,
Brand, Manufacturer
› Store dimension
includes Country,
Continent
Fact tables
› Contains measures (e.g.
Sales Value) and
dimension FK’s
› Dimension columns are
FK’s pointing to the
respective dimensions.

Dimensions & Hierarchies
Hierarchies for the
dimensions are stored
in the dimensional
table itself so there is
no need for the
individual hierarchical
lookup tables be
shown in the model.
Records in dimension
tables correspond to
nouns, the tables are
“short” – 10s to 1,000s
of records
Rich set of attributes,
tables are “wide” –
many columns & the
data changes slowly
Denormalised so no
need to join to further
lookup tables. This
means there is some
redundancy

Fact Tables
Records in fact
tables correspond
to events,
transactions, or
measurements.
Data is added
regularly; the tables
are “long” – often
millions of records
Rich set of
attributes; the
tables are “narrow”
– minimal number
of columns
Low redundancy The most useful
measures are “additive”

Advanced Concepts
Aggregates
Bridge
Tables
Slowly
Changing
Dimensions
Factless
Facts

DATA
MODEL
SOA & XML
messages
BI &DW
Data Lineage
Communicating
with Business
Package
Selection &
Config
Data
Virtualisation
Data Modelling:
It’s NOT just
for DataBase
Design

SOA 101
Enterprise Service Bus
The transport mechanism
System Component
Proves 1 or more services for
the complete architecture
Adapter
Translation where two
worlds collide
Message Queues
Data movement to & from
service components
Message Broker
Policeman directing all traffic
according to the Routing Table

XML Messages Need Data Models
ENTERPRISE SERVICE BUS
DBMS A
System A
DBMS B
System B
XML message

XML versus E/R Structures
Well defined processes for converting
ER/Model into XML
Anyone Remember IMS & DL/1?
› Hierarchical - tree structure.
› Each entity has just one parent.
› Used for transfer of data.
› Shared data appears multiple
times in multiple messages.
XML
› Relational - network structure.
› Each entity can have many parents.
› Used for storage and maintenance
of data.
› Shared data typically appears just
once.
E/R Structure

Data Lineage
SOX LINEAGE
REQUIREMENTS
REPOSITORY BASED
DATA MIGRATION
DESIGN - CONSISTENCY
LEGACY DATA
TAKE ON
SOURCE TO TARGET
MAPPING
REVERSE ENGINEER &
GENERATE ETL
IMPACT ANALYSIS
TRANSFORMATIONS
• What has been done to the
data?
BUSINESS PROCESSES
• Which business processes can be applied to the data?
• What type of actions do those processes perform
(Create, Read, Update, Delete)?
• Audit Trail – who has supplied, accessed, updated,
approved and deleted the data and when?
Which processes have acted on the data?

Data Lineage: e.g. SOX
Financial
reporting &
auditing
requirements
Near real
time reporting
Accuracy of
Financial
statements
Effectiveness
of internal
controls
INFORMATION
MANAGEMENT ESSENTIAL
DATA LINEAGE
IMPLICATIONS
DATA GOVERNANCE
IMPLICATIONS
DATA QUALITY &
DEFINITIONS IMPLICATIONS

Package / ERP systems
Data Requirements For
Configuration & Fit For
Purpose Evaluation
Data Integration &
Governance
Legacy Data Take On Master Data Integration
DATA MODEL

Data Modelling For Packages
/ ERP Systems
Data lineage (particularly important with
Data Lineage & SOX compliance issues)
Master Data alignment
For Data migration / take on
Identifying gaps
For requirements gathering ... But what if
we’ve got to use package X?
CUSTOMER
ORDER
CUSTOMER
ORDER
VS

You Can’t Always Reverse Engineer

Data Virtualisation
Virtual Operational
Data Stores
Shareable Data
Services
D A T A MOD E L
S QL
W E B
S E R VIC ES
S T A R
Virtual
Data Marts
Relational
Views
LEGACY
MAINFRAMES
FILES RDBMS
WEB
SERVICES
PACKAGES
BI, MI AND
REPORTING
CUSTOM APPS
PORTALS &
DASHBOARDS
ENTERPRISE
SEARCH

Communicating With The Business

Data Modelling does not
have to be Complicated!
If you can write a sentence, you can build a data model.
If you understand how your business works, you can build a
data model.
Businesspeople should be involved in the development of data
models, because only they understand the business needs
and rules.
Understanding data modelling basics will help the Business
better communicate with IT

Summary
Data is at the heart
of ALL architecture
disciplines
Data has to be
understood to be
managed
Different levels of
models for different
purposes
It’s NOT just for
DBMS design
Data models are not
(just) art
Professional
development:
certification &
training
All of the Architecture disciplines use the language
(and rules) of the data model

@inforacer
uk.linkedin.com/in/christophermichaelbradley/
+44 7973 184475
infomanagementlifeandpetrol.blogspot.com
Christopher Bradley
I N F O R M A T I O N M A N A G E M E N T S T R A T E G I S T
Chris.Bradley@DMAdvisors.co.uk
D A T A M A N A G E M E N T A D V I S O R S

Data Modelling 101
THANK YOU
Christopher Bradley
Information Strategist

Data modelling 101

More Related Content

What's hot (20)

Viewers also liked (18)

Similar to Data modelling 101 (20)

More from Christopher Bradley (19)

Recently uploaded (20)

Data modelling 101