enterprise taxonomy and schema management: best practices university of washington january, 2006...
Post on 21-Dec-2015
214 views
TRANSCRIPT
Enterprise Taxonomy and Schema Management: Best Practices
University of Washington
January, 2006
Gary CarlsonChief Taxonomist
Company Confidential © 2005 SchemaLogic Inc.
SchemaLogic at a Glance
Company SchemaLogic is a software company based near Seattle, WA, founded in
2001 by ex-Microsoft employees Customers
Includes in-progress production deployments for customers such as Reed Business Information, HP, Church of LDS, Government of UK, and IBM where our products service all 300,000 employees, in more than 80 countries worldwide
Mission To provide both thought and market leadership by delivering the
industry leading Organizational Schema Management software
Classifications (Taxonomies)
Vocabularies (Metadata)
Thesauri
Ontology
Authority Lists
Translations
OrganizationalSchema
Company Confidential © 2005 SchemaLogic Inc.
Product goals…
An easy to deploy, COTS, JAVA web service solution that has been validated by other large organizations
Both a modeling and management platform…the single source of truth for organizational schema
A collaborative framework for change management, with configurable voting parameters
Standards based (XML) import and export Hub and Spoke, Publish and Subscribe synchronization of target
systems Pre-existing adaptors for many current applications Linear scalability up to millions of terms
Company Confidential © 2005 SchemaLogic Inc.
Goal: Centrally Classify the language of the Enterprise
Oracle/
SQL Server/
etc
And Deliver Improved Information Access
“Change Once, Synchronize Many”
Company Confidential © 2005 SchemaLogic Inc.
Successful Management Spans and Integrates the Enterprise
Content management/integration• Taxonomy reconciliation• Standard vocabularies
Data warehouse/reporting teams• Reconcile for BI• Reference data mgt
XML developers• Reuse of XML Schema• Approved changes synchronized
Enterprise Information Architects• Model information flow• Many-to-one mapping, standards
Company Confidential © 2005 SchemaLogic Inc.
Root Problem: Metadata "families“ Reconcilable Differences in Semantics & Structures
Cust_IDCustomer#Client_noCUST1
Severity = 1Sev_Code = A
PRICE:COST:
• Contributor• Author
One Concept: different expressions
BabeBabeOne
Expression: different concepts
Duration = 1Duration = 1
Ambiguity
Company Confidential © 2005 SchemaLogic Inc.
Taxonomies are used in many placesLIST OF COUNTRIES
SITE NAVIGATION RE-USED ACROSS ENTIRE WEBSITE
OTHER TAXONOMY
LINKS
PRODUCTS BY GROUP
PRODUCTS BY BRAND
TWO VIEWS OF THE
PRODUCT TAXONOMY
Company Confidential © 2005 SchemaLogic Inc.
Sharing across the Enterprise
SchemaLogic Enterprise Suite
Numbers & the number system· Ma-0002 Counting· Ma-0003 Integers· Ma-0004 Ordering· Ma-0005 Rounding· Ma-0006 Place value· Ma-0007 Money· Ma-0044 Mental images· Ma-0011 Pi· Ma-0013 Factors· Ma-0014 Prime factors· Ma-0016 Multiples· Ma-8005 Roots· Ma-8004 Powers· Ma-8001 Negative numbers· Ma-8003 Irrational numbers· Ma-8002 Prime numbers· Ma-0018 Standard index form· Ma-5001 Odd & even numbers· Ma-5005 Properties of numbers· Ma-5002 Reading numbers· Ma-5003 Writing numbers· Ma-8026 Number patterns· Fractions, decimals & percentages· Ma-0019 Fractions· Ma-0020 Calculating with fractions· Ma-0021 Equivalent fractions· Ma-0022 Decimals· Ma-0023 Calculating with decimals· Ma-0024 Percentages
Numbers & the number system· Ma-0002 Counting· Ma-0003 Integers· Ma-0004 Ordering· Ma-0005 Rounding· Ma-0006 Place value· Ma-0007 Money· Ma-0044 Mental images· Ma-0011 Pi· Ma-0013 Factors· Ma-0014 Prime factors· Ma-0016 Multiples· Ma-8005 Roots· Ma-8004 Powers· Ma-8001 Negative numbers· Ma-8003 Irrational numbers· Ma-8002 Prime numbers· Ma-0018 Standard index form· Ma-5001 Odd & even numbers· Ma-5005 Properties of numbers· Ma-5002 Reading numbers· Ma-5003 Writing numbers· Ma-8026 Number patterns· Fractions, decimals & percentages· Ma-0019 Fractions· Ma-0020 Calculating with fractions· Ma-0021 Equivalent fractions· Ma-0022 Decimals· Ma-0023 Calculating with decimals· Ma-0024 Percentages
Numbers & the number system· Ma-0002 Counting· Ma-0003 Integers· Ma-0004 Ordering· Ma-0005 Rounding· Ma-0006 Place value· Ma-0007 Money· Ma-0044 Mental images· Ma-0011 Pi· Ma-0013 Factors· Ma-0014 Prime factors· Ma-0016 Multiples· Ma-8005 Roots· Ma-8004 Powers· Ma-8001 Negative numbers· Ma-8003 Irrational numbers· Ma-8002 Prime numbers· Ma-0018 Standard index form· Ma-5001 Odd & even numbers· Ma-5005 Properties of numbers· Ma-5002 Reading numbers· Ma-5003 Writing numbers· Ma-8026 Number patterns· Fractions, decimals & percentages· Ma-0019 Fractions· Ma-0020 Calculating with fractions· Ma-0021 Equivalent fractions· Ma-0022 Decimals· Ma-0023 Calculating with decimals· Ma-0024 Percentages
Numbers & the number system· Ma-0002 Counting· Ma-0003 Integers· Ma-0004 Ordering· Ma-0005 Rounding· Ma-0006 Place value· Ma-0007 Money· Ma-0044 Mental images· Ma-0011 Pi· Ma-0013 Factors· Ma-0014 Prime factors· Ma-0016 Multiples· Ma-8005 Roots· Ma-8004 Powers· Ma-8001 Negative numbers· Ma-8003 Irrational numbers· Ma-8002 Prime numbers· Ma-0018 Standard index form· Ma-5001 Odd & even numbers· Ma-5005 Properties of numbers· Ma-5002 Reading numbers· Ma-5003 Writing numbers· Ma-8026 Number patterns· Fractions, decimals & percentages· Ma-0019 Fractions· Ma-0020 Calculating with fractions· Ma-0021 Equivalent fractions· Ma-0022 Decimals· Ma-0023 Calculating with decimals· Ma-0024 Percentages
Taxonomies and Metadata are consumed by structured and
unstructured information systems
Company Confidential © 2005 SchemaLogic Inc.
A Codification of corporate knowledge and information assets
A way of identifying, maintaining, and updating relationships between different types of information
A tool that can be used to help users get faster access to the most relevant information needed for analysis and decision making
A way of standardizing the way information is described and accessed across an organization
Linking information
A Codification of corporate knowledge and information assets
A way of identifying, maintaining, and updating relationships between different types of information
A tool that can be used to help users get faster access to the most relevant information needed for analysis and decision making
A way of standardizing the way information is described and accessed across an organization
Linking information
What are Taxonomies and Schemas used for?
Company Confidential © 2005 SchemaLogic Inc.
Managing Many Taxonomies
Often have important relationships between them
Integration of taxonomies is similar to integration of systems Allows for re-use, consistency, and utilization
of existing information Requires a higher level of governance and
cooperation between business groups Shines a bright light on business processes
Company Confidential © 2005 SchemaLogic Inc.
Sharing One Taxonomy amongst Many Systems
Sharing a taxonomy has many advantages
Requires flexibility in the modeling tools Requires flexibility from the individual groups
using the taxonomy Requires stronger process for governance Shines a bright light on business processes
Company Confidential © 2005 SchemaLogic Inc.
Three faces of the solution
Component Role
Modeling Environment Must provide flexible and robust features for the meta-model.
Must allow for multiple views of the meta-model
Change Control (workflow)
Must provide impact analysis so the cost of change can be determined prior to change
Must allow for full notification to stakeholders
Synchronization (import/export of metadata definitions)
Flexible to connect to many systems
XML, JDBC, Legacy systems, etc.
Full import and export of definitions
Company Confidential © 2005 SchemaLogic Inc.
Three primary constituencies
Stakeholders TasksBusiness Owners – Product Managers, Report producers, Legal,HR, etc
Drive particular business requirements for metadata initiatives. Represent many different business units
Metadata Stewards – Taxonomists, Library Science Professionals
Create meta-model which supports the business requirements of all business users and IT requirements. Need to work across business units.
IT - Data Architects, System Integrators,
Provide input into the feasibility and cost of making changes. Provide essential background knowledge into existing models.
Implement changes to the meta-model.
Represent many different enterprise systems
Company Confidential © 2005 SchemaLogic Inc.
The Fourth Constituency…
Organization Executives…
Need a complete view of the customerSales reports by and across regionsLegal auditing requirementsEtc
Company Confidential © 2005 SchemaLogic Inc.
And the Fifth
The Customer…
Consistent informationGood customer supportGood on-line experienceEtc…
Company Confidential © 2005 SchemaLogic Inc.
Re-visit the Business Requirements
Component Stakeholders
Modeling Environment
Where is the metadata managed
Business Owners What metadata is needed
Who needs the metadata
Why is it needed
Change Control
How is it modified
Who is involved
Metadata Stewards
Who manages the overall structure
Synchronization
When/Where is it used
IT How is it implemented
What systems are involved
What level of integration is required
Company Confidential © 2005 SchemaLogic Inc.
The Crux of the matter
A good Taxonomy analysis is really just a good business analysis.
All components should map to a direct business requirement.
A good metadata solution may (and will) comprise far more than metadata
Company Confidential © 2005 SchemaLogic Inc.
SchemaLogic Solution Architecture
.SLV XML*
Model
Investigate
Administer
ManageChange
Search
Auto-cat&
ContentMgmnt
Portal
Reports
CustomApps
Classifications
Taxo
nom
ies
Voc
abul
arie
s
Authority Lists
Onthology Th
esau
rus
Subscribe
Syn
chro
nize
SchemaLogic Product Suite
INTEGRATE
TRANSFORM
Company Confidential © 2005 SchemaLogic Inc.
SchemaServer SDK
Workshop Catalog ActiveTerms Custom UI
SchemaLogic Desktop Applications
SchemaLogic Product Architecture
IntegrationServices
ModelingServices
Collaboration
ServicesImportExport
Business Rules
Security &Permissions
DeltaUpdates
Full Thesaurus
Impact Analysis
TargetedSystems
User/SystemViews
Versioning
REPOSITORY SERVICES
SchemaServer
J2EE
AIX/HP-UX/Solaris/ Win-2003
DB2/SQL/Oracle
Inte
gra
tor
Fra
me
wo
rk
Adaptors
Stellent
Documentum
Vignette
Autonomy
Verity
BEA
IBM WebSphere
IBM OmniFind
Company Confidential © 2005 SchemaLogic Inc.
SchemaLogic® Workshop™
Allows all the constituencies to work in one place
Primary UI Role-based behavior Plug-in architecture Large vocabulary
manipulation Java WebStart
Company Confidential © 2005 SchemaLogic Inc.
SchemaServer
SchemaLogic® Integrator™ and Adaptors
Integrator #1
Adaptor #1Integrator UI
Workshop
Adapters Access in Workshop UI Binds to Target Systems Perform Synchronization File Transfer Integrations
Integrator Servlet-based Web Service Lightweight, highly scalable Logging/Versioning/Audit Trail Reduces Adaptor Complexity
Synchronize Adaptor #2
Integrator #2
Integrator #3
Synchronize
Synchronize
Hub and Spoke, Publish and Subscribe Schema Synchronizations
Company Confidential © 2005 SchemaLogic Inc.
Example: SchemaLogic SharePoint Adaptors
Benefits
Site Governance Site Consistency
Rich Thesaurus Enhanced Search
SchemaLogic
SharePointSite Server
SharePointTemplate Adaptor
SharePointThesaurus Adaptor
MS Search
MSThesaurus
SiteTemplates
Template Synchronization
Thesaurus
Synchronization
SchemaLogicIntegrator
Tem
plat
e
Synch
roni
zatio
n
Index and Search
(StructuredLists)
Company Confidential © 2005 SchemaLogic Inc.
The SchemaLogic Taxonomy Solution
Taxonomies and Classification Models imported from source systems and synchronized into enterprise taxonomy.
SchemaLogic enables global terminology to be centrally managed across distributed teams and consuming systems.
Standard vocabulariesare used for document creation which establishes accurate semantic context of information.
Standard classifications applied to all content silos ensure content-store metadata is consistent and complete.
Global expertise is better leveraged via greater ease of portal and content navigation.
Search engines are aware of information context which improves search-result relevance.
Company Confidential © 2005 SchemaLogic Inc.
SchemaLogic® Integrator™
The SchemaLogic integration framework
Adaptors for target systems
Adhoc or scheduled synchronizations
SchemaServer(Content Class) (Vocabulary) (Vocabulary)(Content Class)
ProductsCustomerGeographyDublinCore
Portal
Constant Metadata
Structures
ValueLists
Taxonomies
CMS
Constant Metadata
Structures
ValueLists
Auto-categorizationand Search
Taxonomyof Categories
Browseand
Navigation
Binding
Subscriptions
Adapter: CMS
Binding
Subscriptions
Binding
Subscriptions
Binding
Subscriptions
Binding
Subscriptions
Adapter: Auto-Categorization and Search Adapter: Portal
Inte
gra
tor
Se
rve
r
Manage Bindings
and Subscriptions
Manage Bindings
and Subscriptions
Company Confidential © 2005 SchemaLogic Inc.
SchemaLogicSchemaServer
ExpertiseTaxonomy
● Skills● Roles
On Demand Workplace
“In 2005, SchemaLogic will enable management of the Expertise and Enterprise Taxonomies that support Bluepages and the On Demand Workplace.”
OmniFind Search
Corporate.com
XML Repository
User Reports
Bulk Import
Manual Edits
Bulk Update
(Translations)
“Marketplace is helping consultants find niche skills and get to work on contracts faster, factors that could help boost revenue, not just cut costs.” ― Wall Street Journal, August 2005
Customer Scenario 2: Enterprise Taxonomy Management for Global Employee Portal
Company Confidential © 2005 SchemaLogic Inc.
SchemaLogicSchemaServer
Microsoft SharePoint Controlled
Vocabularies
Document profilesList metadataSite templates
SynonymsFacet ExpansionAbbreviationsLocalizationMisspellings
Centralized Governance, Thesaurus and Search forWindows SharePoint Services and SharePoint Portal Server
1000+ SharePoint Portal Sites
Benefit:Enhanced Findability
and Governance
Customer Scenario 3:Global Pharmaceutical Company
Company Confidential © 2005 SchemaLogic Inc.
Search Strategy – 12MM Member Church
DynamicDMAM
(Telescope/Virage)
INDEXES
Index to Periodicals
METADATA
Static, Authoritative
Additional Indexes
Index
Topical Guide
Bible Dictionary
Guide to Scriptures
DICOM (Documentum)
WCM(Vignette)
Other
Content Reference/
Article Reference
“See”“See also”
Terms/Nested Terms
Monitor metadata
Provide access for metadata
for manual assignment
Mine existing date to:Adjust taxonomies
Assist in building dynamic indexesAutomatically assign metadata
Provide concept clusters under specified
taxonomy points; Captured searches; Ontology capability
SchemaLogic
Enterprise Taxonomies