DAMA APRIL 18, 2000 HIGHLIGHTS FROM: DAMA INTERNATIONAL SYMPOSIUM METADATA CONFERENCE DAMA PRESIDENT

1
DAMAAPRIL 18, 2000HIGHLIGHTS FROM DAMA
INTERNATIONAL SYMPOSIUM/META-DATA
CONFERENCEDAMA PRESIDENTS COUNCILPRESENTED
BYJennifer Cooke, Systems Analyst/Developer at
Hewlett PackardChimei Shea, Data Analyst at The
Standard InsuranceRita Wheeler, Data Architect
at The Standard InsuranceBrenda Finley, Data
Analyst at The Standard InsuranceMorgan Goodwin,
Data Analyst at The Standard InsuranceHeidi
Babcock, IT Engineer at Hewlett Packard Martha
Shehorn, IT Engineer at Hewlett Packard
2
AGENDA

• DAMA I President/Portland Chapter Liaison Mike
  Brackett
• DAMA Presidents Council Meeting Rita Wheeler
• DAMA I Symposium/Metadata Conference
  Highlights Chapter Members
• DAMA I Symposium/Metadata Conference 2001 Rita
  Wheeler

3
DAMA I President andPortland Chapter
LiaisonMike Brackett

• DAMA Vision for 2000
• Function of DAMA I
• Role of DAMA I Liaison

4
DAMA Advisors

• John Zachman
• Peter Aiken
• Clive Finkelstein
• Ron Powell

5
DAMA President's Council Meeting March 19,
2000Objectives

• Identify and Confirm Major Items for Follow-up
• Identify Follow-up Committees

6
DAMA President's Council Meeting March 19,
2000Committees

• Membership
• Education
• Central Address Maintenance
• Website

7
DAMA I Symposium/Metadata Conference Highlights

• Metadata Management in a CIF - Jennifer Cooke
• Questions Metadata Can Answer - Chimei Shea
• Information Models for Metadata - Rita Wheeler
• Understanding DW Strategically - Brenda Finley
• Enterprise Application Integration (EAI) - Rita
  Wheeler
• Data Quality - Heidi Babcock
• Data Stewardship - Heidi Babcock
• Graeme Simsions Top 10 List - Martha Shehorn

8
Metadata Management in a CIF
Jennifer CookeHewlett Packard - Corvallis,
ORSystems Analyst/Developer

• Taken from
• The Top Issues in Implementing a Data
• Warehouse and Corporate Information Factory
• by Claudia Imhoff

9
Metadata Management in a CIF CIF Poster

• http//www.dmreview.com/posters/cif/

10
Metadata Management in a CIF Metadata Management

• Provides the necessary details to promote data
  legibility, use and administration. Contents are
  described in terms of data about data, activity
  and knowledge.
• Improves data use and re-use, transfers
  knowledge, and reduces cost for information
  requirements.

11
Metadata Management in a CIF Metadata Repository

• Maintains data about data
• Definitions
• Alias names
• Standard calculations
• Who uses the data?
• Where is it now? And where did it come from?
• An important formal component of the Corporate
  Information Factory
• Provides data legibility to the end user
• Provides administrators information to manage the
  environment

12
Metadata Management in a CIF Three Categories
of Metadata

• Technical Metadata
• Construction of the warehouse
• On-going upkeep
• Business Metadata
• Business view
• Analyst user interest
• Generally manually captured and maintained
• Administrative Metadata
• Health and well-being of CIF

13
Metadata Management in a CIF Types of Metadata

• Shared Metadata
• Satisfies standards
• Resolution of integration issues
• Accessed in same manner by everyone
• Easier to understand
• Improved decision making and communication

• Local Metadata
• Need to know basis
• Personnel information
• Sales information
• Not of general interest
• Customer information

14
Metadata Management in a CIF Metadata
Architecture

• Buy a central repository
• Buy a metadata management tool
• Build your own version

15
Metadata Management in a CIF Centralized /
Homogeneous Architecture

• PROS
• Enforce enterprise standards
• Records metadata from all sources
• Integration and shareability
• Access is the same for everyone

• CONS
• Can be a bottle neck
• Resource intensive to maintain
• Redundancy
• Source of record in tools, not repository
• No local metadata accommodation
• Nothing today to automated feeds to sources

16
Metadata Management in a CIF Decentralized /
Heterogeneous Architecture

• PROS
• Shared and local metadata
• Diverse technology for capture and access
• Reduces redundancy
• Source of record where metadata is captured

• CONS
• Synchronization issues between metadata databases
• Difficult to declare THE source
• Not many tools available today

17
Metadata Management in a CIF Start Now!

• Metadata Strategy Document
• Create a metadata model
• Declare contents of metadata types
• Capture and delivery mechanisms
• Maintenance guidelines
• Pick a product or solution and maintain it
• Create an effective strategy and direction for a
  two year period

18
Metadata Management in a CIF Design Tips

• Build your architecture gradually
• Integrate business and technical metadata
• Deliver via the Intranet/Internet
• Educate, sell to, and build for the end user
• Will grow in importance and evolve
• Allow for new metadata requirements
• Standards and limited update authority

19
Metadata Management in a CIF Vendors

• Centralized
• CA (Platinum)
• PR/MVS
• PR/OEE
• OIM on other platforms
• One Meaning - Marlow
• Microsoft OIM on SQL Server
• Viasoft - Rochade

• Decentralized
• Pine Cone Systems - Meta Exchange Manager
• VIT (Access, less true management)
• Ardent - Meta Stage
• Sybase - Warehouse Control Center

20
Metadata Management in a CIF Want to Learn
More?!?

• Building the Operational Data Store - Claudia
  Imhoff and Bill Inmon
• The Corporate Information Factory - Claudia
  Imhoff and Bill Inmon
• Are you an Inny or Outty? - Claudia Imhoff, DM
  Review, September 1999.
• The Corporate Information Factory - Claudia
  Imhoff, DM Review, December 1999
• Exploration Warehouse The Final Frontier -
  Claudia Imhoff, DM Review, February 2000.

21
Questions Metadata Can Answer
Chimei SheaThe Standard Insurance - Portland,
ORData Analyst

• Taken from
• Questions Metadata Can Answer
• by Robert Seiner

22
Questions Metadata Can Answer In Your Current
Environment

• Can my company answer these questions?
• What is it costing my company to answer these
  questions?
• What is the result when we arent able to answer
  these questions?
• How will we benefit from being able to answer
  these questions effectively and efficiently?

23
Questions Metadata Can Answer In Your Current
Environment

• Are the answers to these questions important?
• Are the answer to these questions always
  available?
• Will the IT division perform better if they
  have ready access to this information?
• Cost savings and competitive advantage are
  associated with managing data through metadata.

24
Questions Metadata Can Answer Metadata Question
Categories

• Database
• Data Model
• Data Movement
• Business Rule
• Data Stewardship

• Application Component
• Data Access / Reporting
• Rationalization
• Data Quality
• Computer Operations

25
Questions Metadata Can Answer In Your Current
Environment

• See Handout for all the questions

26
Information Models for Metadata
Rita WheelerThe Standard Insurance - Portland,
ORData Architect
Taken from Multiple Sessions
27
Information Models for Metadata Two Standards

• OIM - Open Information Model
• MDC - Metadata Coalition
• CWM - Common Warehouse Meta Model
• OMG - Object Management Group

28
Information Models for Metadata OIM - Open
Information Model

• Effort began by Microsoft and partners in
  October, 1996
• Developed with over 20 partners and reviewed by
  over 300 companies
• Goal enable sharing and reuse of metadata by
  providing definitions of core types
• Accommodates vendor and tool specific extensions
• Standardized by MDC in July, 1999

29
Information Models for Metadata MDC - Metadata
Coalition

• Microsoft transferred rights (to evolve the OIM)
  to the MDC in Dec. 1998
• MDCs goal to define a technology independent
  and vendor-neutral metadata standard.
• The MDC will
• use UML (Unified Modeling Language) as the
  specification language
• use XML as interchange format for metadata
  described by the OIM
• work with other standards organizations (OMG) to
  promote common metadata models

30
Information Models for Metadata OIM Subject
Areas

• Analysis and Design
• UML
• Common Data Types
• Database and Warehousing
• Relational Schemas
• Data Transformations
• Multidimensional Schemas
• Legacy Databases
• Component Management
• Knowledge Management

31
Information Models for Metadata CWM - Common
Warehouse Meta Model

• Co-submitting Companies IBM, Unisys, NCR,
  Hyperion, Oracle, UBS, Genesis, Dimension EDI
• 7 Supporting companies
• Goal metadata management and integration for
  Data Warehousing/Business Intelligence
• UML as the standard language for defining models
  of metadata

32
Information Models for Metadata OMG - Object
Management Group

• Initial submission on 9/17/1999
• Final submission on 2/11/2000
• Adopted specification targeted 3/2000 or 6/2000
• Available specification targeted 9/2000 or 12/2000

33
Understanding DW Strategically
Brenda FinleyThe Standard Insurance - Portland,
ORData Analyst
Taken from Understanding Data Warehousing
Strategically by Bernie Boar
34
Understanding DW StrategicallyTypical Data
Warehouse Justifications

• Improved Decision Making
• Empowerment
• Leveraging of Operational Data
• Customer Intimacy
• Process Control
• Customer, Industry Self Knowledge
• But what is the compelling strategic logic
• of data warehousing?

35
Understanding DW StrategicallyStrategy Views

• Academic View
• Provide direction, Concentration of Effort,
  Constancy of Purpose and Flexibility as Business
  moves to improve position in all strategic areas
  
• Pragmatic View
• Find a way short (the shorter, the better) of
  BRUTE FORCE to accomplish ones ends.
• Business View
• Struggle for Advantage. One with more advantages
  wins, the one with fewer loses. The purpose of
  strategy is the building and sustaining of
  advantage.

36
Understanding DW StrategicallyStrategy Views

• Dimensions of Advantage
• Cost
• Differentiation
• Focus
• Execution
• Maneuverability

The Culmination of Advantage Sustainable
Competitive Advantage
37
Understanding DW StrategicallyOther Thoughts .
. .

• Strategic Thinking is 3-Dimensional
• Time Past, Present, Future
• Substance Abstract, Concrete
• Cardinality Number of Concurrent Issues

Data Warehousing permits the thinking across time
- A critical factor for a strategist

• Rising Tide Strategy is Powerful due to Leverage
• Reuse, Sharing, Duplicating, Layering,
  Multiplying
• Leverage Value (Individual Payoff x Instances)

Get as many people on data warehouse as possible
(rising tide), this increases leverage
38
Understanding DW StrategicallyStrategic Paradox

• Strategy developed and executed within a context
  of extreme conflict, intelligent reactions,
  counter measures
• Linear Logic - Common Sense
• Good for daily activities
• Uses Inductive / Deductive Reasoning
• Paradoxical Logic - Reversal of Opposites
• Necessary for strategic decisions
• Holding on to an advantage too long becomes a
  disadvantage (too much of a good thing)
• Do the reverse of what linear logic would dictate
• For example To empower, there must be strong
  leadership. To concentrate power, distribute
  it.
• Turn exceptional features into the norm

39
Understanding DW StrategicallyStrategic Paradox

• Information Age
• Speed is of the essence
• Global Information Exchange
• Destruction of Barriers to Entry
• Hyper Competition
• IT Fighting
• Data Warehousing at the edge of conflict
• Subject to strategic paradox
• Requiring strategic logic

Data Warehousing is most successful when one
does not seek to use just enough means, but an
excess of means. Data Warehousing is most
efficient when it is applied in excess.
40
Understanding DW StrategicallyManeuver

• Disrupt their OODA by Maneuvering Observe Orient
  Design Act

GO FORTH WHERE THEY DO NOT EXPECT IT, ATTACK
WHERE THEY ARE UNPREPARED. SUN TZU - THE ART OF
WAR
Destroy the ability of the opponent to
efficiently or effectively execute this cycle.
41
Understanding DW StrategicallyManeuver
But to maneuver, requires knowledge - the
foundation of knowledge is the DATA
WAREHOUSE SUMMARY Data Warehousing is a
mandatory prerequisite to engage in a maneuver
market strategy.
42
Understanding DW StrategicallyData Warehousing
Strategic Intent
The strategic intent of our data warehousing
strategy is to enable the business to win in the
marketplace every day with every customer and
with every purchase. By repositioning our
operational data and combining it with selected
foreign data, we will empower our employees so
that they can routinely delight and excite our
customers. Through our unique appreciation of
the value of our data assets, we will elevate our
data warehouses to the point where they become a
compelling and durable contributor to the
sustainable competitive advantage of the
business. In this way, data warehousing will
enable the business to impress its attitude on
the marketplace and prevail over its competitors
who have already lost.
43
Data Quality
Heidi BabcockHewlett Packard - Corvallis,
OREnterprise Reference Data Manager
Taken from Multiple Sessions
44
Data Quality

• We need to be Quality people, not clean up
  people. - Larry English
• Pre - Act vs. React
• Adding value back to the data and underlying
  structure

45
Data Quality Problem - Michael Bracket

• Redundancy
• average redundancy factor of 10
• Variability
• average variability factor of 15-20
• Disparity Cycle
• cant find-trust-access data
• independent, siloed solutions
• no standards, documentation, or integration

46
Data Quality Best Practices

• Avoid Suck and Squirt Approaches - Michael
  Bracket
• similar to spatula method
• need to integrate
• Behavior driven - Larry English
• fix the immediate problem
• investigate and correct root cause
• The Business Manager must be accountable for
  data quality - Larry English
• It is 10x more expensive to clean up data
  downstream than it does at the source - Larry
  English
• Make no defects - Pass on no defects - Accept
  no defects - Larry English
• Clean up old, expired, obsolete data - Larry
  English

47
Data Quality Principles - Larry English

• Customer focus
• market focus
• cust satisfaction
• partnership
• Process Improvement
• process definition
• product specification
• team work
• continuous process improvement and business
  process re-engineering
• Scientific Methods
• statistical quality control
• PDCS

48
Data Standards Best Practices

• When two or more standards conflict, there is
  not standard - Michael Bracket
• Standards must be developed within a common
  data architecture for high-quality data - Michael
  Bracket
• If a Standard isnt followed, there is no
  standard - Larry English

49
Data Value - Larry English

• Every time data is moved from one system to
  another, it is worth less
• loss of meaning during translation
• cost of translation and moving
• Exploit data retrieval for highest ROI
• Re-Use is a major indicator of quality and
  value
• Unjustified and uncontrolled redundancy is a
  major indicator of nonquality (low value)

50
Data Content Quality Measures - Larry English

• Completeness
• Validity
• Accuracy
• Precision
• Non-duplication
• Consistency
• Timeliness
• Rightness

51
Data Content Quality Measures - Larry English

• Determine the baseline
• what should the data quality look like
• Obtain a maintainable, achievable target over a
  period of time
• Use audit marking
• suspect
• incorrect
• unreliable
• corrected

52
Quick Wins for Info Quality - Larry English

• Quantify cost of quality problems
• Measure and report to info producers
• Develop SLA between business areas
• Put info quality on every meeting agenda

53
Data Stewardship
Heidi BabcockHewlett Packard - Corvallis,
OREnterprise Reference Data Manager
Taken from Multiple Sessions
54
Data Stewardship

• Multi-leveled
• Ownership is w/in Business

55
Data Stewardship - Michael Bracket

• Strategic Level
• executive level
• major data subjects
• Tactical Level
• data liaisons between exec level and knowledge
  worker and between stakeholders
• Detail Level
• knowledge worker
• SDA focus
• architectural responsibility
• may cross org lines

56
Data Stewardship RR - Burt Parker

• Data Policy Steward
• exec level
• governance circle
• Enterprise Data Steward
• pan organization
• may be segmented per business entity
• Business Data Steward
• vertical business view
• aligned with Ent DS
• Project Data Steward
• application view
• aligned with Bus DS
• Operational Data Steward
• quality input and/or use
• aligned with Bus DS

57
Conference Take-Aways

• Reafirmation
• We are all in the same boat
• The boat is gaining water
• Who is fixing the leak?
• Who has the oars?

• Sensibility
• Tweener approach
• Must be able to
• . Meet the business need
• AND
• . Add value back
• New and sustaining investments
• Forward movement
• AND
• Keeping it alive

58
Great Quotes

• Tools and Technology do not understand the
  data. Only people understand the data. - Michael
  Bracket
• If the foundation of the house isnt square and
  level, you fight it at every level up to the last
  shingle on the roof. - Michael Bracket
• We must target our audience and resist
  expanding our IT Knowledge. - Michael Bracket
• Data left alone will deteriorate and depreciate
  over time. - Linda Paliaro

59
Great Quotes

• It takes 1/5 the resources to establish Data
  Quality as it does to scrap and rework. - Larry
  English
• Go Slow so you can Go Fast - Japanese saying
• We need to be Quality people, not clean up
  people. - Larry English
• Vertical, industrial-age mindsets beget poor
  quality systems and data. - Larry English
• Cleansing the data addresses the symptom. It
  doesnt address the root cause. - Larry English

60
EAI - Enterprise Application Integration
Rita WheelerThe Standard Insurance - Portland,
ORData Architect
Taken from Multiple Sessions
61
EAI - Enterprise Application IntegrationHandouts
- Articles

• One Customer View - Dan Vander Hey - Intelligent
  Enterprise
• EAI Directions - Nelson King - Intelligent
  Enterprise
• Understanding Data Level EAI - Andre Yee

62
Graeme Simsions Top 10 List
Martha ShehornHewlett Packard - Vancouver, WAIT
Engineer
Taken from Closing Keynote Presentation Data
Management - Where to from Here? by Graeme Simsion
63
Graeme Simsions Top 10 List

• Our goals the business goals
• Dont take on the responsibility for identifying
  value for the rest of the business let them do
  it
• Focus less is more
• Dont get it right get it done
• Look outside for standards
• Do it (whatever) with the business goal in mind
• Review the skills base
• Measure the benefits
• Dont build infrastructure first build it as you
  go
• Stay sane

64
Reference Web Sites

• www.dama.org
• www.dbdsolutions.com (Adrienne Tannenbaum)
• www.kpiusa.com (Knowledge Partners Inc.)
• www.tdan.com (The Data Administration Newsletter)
• www.zifa.com (Zachman Institute for Framework
  Advancement)
• www.infoimpact.com (Larry English - Information
  Impact)
• http//fast.to/peteraiken/ (Peter Aiken, DAMA
  Advisor)
• www.BRSolutions.com (Ron Ross - Business Rule
  Solutions)

65
Reference Web Sites

• www.guide.org
• www.infoadvisors.com
• www.eGroups.com (Data Management List)
• www.inconcept.com (Journal of Conceptual
  Modeling)
• www.IntelSols.com (Intelligent Solutions -
  Claudia Imhoff)
• www.essentialstrategies.com (David Hays)
• www.ebgconsulting.com (Ellen Gottesdiener)
• www.ies.aust.com (Information Engineering
  Services - Clive Finkelstein)

66
Reference Web Sites

• www.dmreview.com (DM Review)
• www.intelligententerprise.com (new free magazine)
• www.mdcinfo.com (Meta Data Coalition)
• http//msdn.Microsoft.com/repository/
  (Microsoft's Repository)
• Repository News Group msnews.microsoft.com,
  micrsoft.public.repository
• www.omg.org (Object Management Group)
• www.omg.org/techprocess/meetings/schedule/CWMI_RF
  P.html (OMG CWMI RFP web page)
• www.cwmforum.org/ (CWM Forum web Site)

67
Reference Web Sites

• www.datawarehouse.com
• www.dw-institute.com
• www.km.org (Knowledge Management Consortium
  International)
• www.dataware.com (Dataware Technologies)
• www.egltd.com (Enterprise Group - David Marco)
• www.w3.org (World Wide Web consortium)
• www.xml.org
• www.xml.com
• www.biztalk.org

68
Reference Books

• Building Corporate Portals with XML - Clive
  Finkelstein and Peter Aiken
• Building and Managing the Meta Data Repository -
  David Marco
• Data Modeling Essentials - Graeme Simsion
• Data Model Patterns - David Hay
• The Data Model Resource Book - Len Silverston,
  Bill Inmon, Kent Graziano
• Implementing a Corporate Repository - Adrienne
  Tannenbaum

69
Reference Books

• Improving Data Warehouse and Business
  Information Quality - Larry English
• Handbook of Relational Database Design - Barbara
  von Halle
• The Handbook of Data Management - Barbara von
  Halle
• The Data Modeling Handbook - Michael Reingruber
  William Gregory
• A Practical Guide to Logical Data Modeling -
  George Tillmann

70
Reference Books

• Building Quality Databases with IDEF1X - Thomas
  Bruce
• Business Rule Concepts - Ron Ross
• Business Rule Book - Ron Ross
• Data Sharing using a Common Data Architecture -
  Mike Brackett
• The Data Warehouse Challenge Taming Data Chaos -
  Mike Brackett
• The Art of War - SUN TZU (strategy)
• Practical Steps to Aligning IT with Business
  Strategies - Bernie Boar

71
Reference Books

• Data Warehouse from Architecture to
  Implementation - Barry Devlin
• Managing the Data Warehouse - Inmon, Welch and
  Glassey
• Data Warehousing the Route to Mass Customization
  - Sean Kelly
• Data Warehousing Concepts, Technologies,
  Implementations, and Management - Harry Singh
• Data Reverse Engineering Slaying the Legacy
  Dragon - Peter Aiken

72
The DAMA INTERNATIONAL and METADATA
CONFERENCE A DATA ODYSSEY
March 5, 2001 to March 8, 2001
ANAHEIM, CALIFORNIA


Disney and DATA!!
WHAT MORE COULD YOU WISH FOR