Managing Namespaces in XML With Metadata Registries

1
Managing Namespaces in XML With Metadata
Registries

• Judith Newton
• NIST
• jnewton_at_nist.gov
• XML Registry Project Team Meeting,
• May 21, 2003

2
Agenda

• Namespaces as metadata
• Introduction to an ISO 11179-based metadata
  registry (MDR) model
• Naming in the MDR
• How an MDR can complement an XML registry

3
What is Metadata?

• A type of data that describes and defines other
  data
• Content, quality and condition
• Who owns it? How was it collected? How current is
  it?
• How many names is it known by?

4
Functional Namespaces

• Traditionally, data managers have divided up an
  enterprises metadata by subject area or
  stewardship domain
• Business rules have been applied to these
  namespaces
• Intermixing happens Person metadata used by
  Human Resources, Customer Service, etc.

5
Previous External Namespace Arrangements

• One-to-one agreements
• Small groups of like-minded folks
• By fiat, like Wal-mart

6
External Namespaces

• Communities of Interest are driving creation of
  namespaces
• Common schema development for subject areas
• Examples ebXML (OASIS), mathML, eCommerce
  (BizTalk)

7
Namespaces From Information Architecture
8
Data Model
LOCATION
PERSON
ADDRESS
GEOGRAPHIC
EMPLOYEE
CUSTOMER
Latitude Longitude Altitude
Street County City Country
Name Position
Name Order History
9
Good Practices for Namespace Management Tools

• Data Model
• Classification of Terms Ontology, taxonomy,
  glossary, keywords
• Metadata Repository or Registry
• XML Schemas

10
Good Practices for Namespace Management Process

• Appoint data stewards
• Publish the semantic content of tools
• Build bridges
• Specify change mechanisms

11
Good Practices for Namespace Management Goals

• Understandability - Users should readily
  comprehend the data and metadata, both
  structurally and semantically
• Trustworthiness - The metadata should document
  the source, collection method, security level,
  and other factors so users can assess the
  authority of the data
• Interoperability - Enough information must be
  provided to users concerning interfaces, whether
  formalized or ad-hoc (or just potential)

12
Example of Namespace Derivation Structure
Proposed scheme for establishing namespaces for
the U.S. government Namespaces organized by
subsetting the dot-gov domain into smaller units
by agency, for example   urnusgovgsa   In a
schema, gsa would be identified by a namespace
prefix   lt?xml version1.0 encodingUTF-8?gt lt
xsdschema xmlnsxsdhttp//www.w3.org/2001/XMLSc
hema xmlnsgsaurnusgovgsa
13
Example of Namespace Prefix in an XML Instance
Document
Examples of instances qualified by the "gsa"
prefix   ltgsaAgencyNamegtGSAlt/gsaAgencyNamegt ltgs
aAgencyIDgt9986lt/gsaAgencyIDgt ltgsaContactPartyID
gt222345897lt/gsaContactPartyIDgt ltgsaOrderedQuanti
tygt100lt/gsaOrderedQuantitygt ltgsaOrderedQuantityA
mountgt399lt/gsaOrderedQuantityAmountgt    
14
XML Technical Namespace Definition
The XML Recommendation defines a Namespace as a
conceptual space in which all names are
unique. Namespace becomes a mechanism to
distinguish among meanings of terms, for instance
bar in the legal sense and bar in the
fence-manufacturing community. In an XML Schema,
an IRI is used to define the namespace. The IRI
may not resolve to any actual content.
15
Agenda

• Namespaces as metadata
• Introduction to an ISO 11179-based metadata
  registry (MDR) model
• Naming in the MDR
• How an MDR can complement an XML registry

16
SC 32 Data Management and
Interchange Meta Data Registry
Projects ISO/IEC 11179 Metadata Registries, Parts
1-6 Framework Classification Registry
metamodel and basic attributes Formulation of
data definitions Naming identification
principles Registration ISO/IEC TR 20943-1,
Procedures for achieving MDR content
consistency-Data elements to be published any day
now ISO/IEC TR 15452, Specification of Data Value
Domains, published in 2000. To be replaced by
ISO/IEC FDTR 20943-3, Procedures for achieving
MDR content consistency- Value Domains ISO/IEC
20944 series APIs and other interfaces
17
High-level Metamodel
Conceptual_Domain
Data_Element_Concept
Value_Domain
Data_Element
18
Administered Item Naming Example

Conceptual_Domain Name Country
Data_Element_Concept NameCountry Identifier
Value_Domain Name English-Language
Country Short Name
Data_Element
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
XML Name
23
(No Transcript)
24
Agenda

• Namespaces as metadata
• Introduction to an ISO 11179-based metadata
  registry (MDR) model
• Naming in the MDR
• How an MDR can complement an XML registry

25
Names

• Names are associated with contexts within the MDR
• Namespaces may be contexts or classifications
• XML contexts may be namespaces, DTDs, schemas or
  other structures
• Other context examples include application
  systems, programming languages, report headers

26
Structured Names

• Scope
• Authority
• Semantic rules
• Syntactic rules
• Lexical rules
• Uniqueness rule

27
Semantic Content of Names

• Semantic rules enable meaning to be conveyed
• Principle Name parts are meaningful to users
• Name parts are derived from structure sets that
  identify relationships among (classify) members

28
Syntax of Names

• Syntax specifies the arrangement of parts
  within a name
• Principle The name parts are arranged in a
  specified order
• The arrangement may be relative or absolute, or
  some combination of the two

29
Lexical Considerations of Names

• These concern the preferred and derogated
  terms, synonyms, abbreviations, part length,
  spelling, character set, case sensitivity, etc.
• Principle All names governed by a specific
  naming convention have a consistent appearance
• Control word form and vocabulary

30
Agenda

• Namespaces as metadata
• Introduction to an ISO 11179-based metadata
  registry (MDR) model
• Naming in the MDR
• How an MDR can complement an XML registry

31
Lexical Rules for XML Context

• Lexical rules
• Nouns are used in singular form only. Verbs (if
  any) are in the present tense.
• Name parts are separated by capitalizing the
  first character of the second thru nth word
• All words in the name are in mixed case.
• Abbreviations, acronyms, and initialisms are
  allowed.
• Example CountryMailingAddressName

32
11179 - XML Collaboration
lt?xml version"1.0"?gt ltshipTo
gt ltnamegtAlice Wilsonlt/namegt
ltstreetgt161 North
Streetlt/streetgt ltcitygtHappy
Valleylt/citygt
ltstategtMOlt/stategt
ltzipgt63105lt/zipgt ltcountry
codegtUSAlt/country codegt
lt/shipTogt
33
The Need for XML Name Management

• XML provides a common information exchange format
  without common semantics
• Each community of interest is defining their own
  namespaces
• How can the process be standardized?
• How can we address the data warehouse problem of
  semantic integrity?

34
XML Name Management Using MDR Registries

• An MDR metadata registry can link the XML syntax
  to common semantic content
• XML namespaces can be defined as contexts or
  classifications within MDR registries
• Names can be defined within and among namespaces
• Naming convention principles can be applied to
  XML names

35
XML Name Management Using 11179 Naming Conventions

• Part 5 of ISO/IEC 11179 addresses naming
  conventions developed using a set of principles
• These principles can be applied to XML names
• ebXML has already produced a draft standardebXML
  Core Components Dictionary Entry Naming
  Conventions at
• http//www.ebxml.org/specs/index.htm

36
Summary

• XML names defined as contexts in MDR
• IDI links XML schema objects to MDR registry
  entries (not just for names but other semantic
  information as well)
• XML naming conventions defined in MDR registry
• Adapt 11179-5 naming principles to XML naming
  conventions

37
The final version of 11179-32003 is now
available to download. 1. Go to www.jtc1.org 2.
Select Freely Available Standards at the bottom
of the left column 3. In LiveLink Go to the
Public areas 4. Scroll to 11179-3