Welcome and Introduction

1
Welcome
1
K 1
Bruce Bargmeyer
Lawrence Berkeley National Laboratory University
of California Tel [phone],
bebargmeyer_at_lbl.gov
2
Welcome

• Welcome
• Conference Logistics
• Internet connections SSID and password in your
  packets.
• Meals
• Breakfast with room
• Lunch 90 minutes
• Reception Tonight at 1930 in Emerald Room
• Agenda
• Prompt start and stop of presentations
• Some late breaking changes
• Thanks to Program Committee
• Professor Hajime Horiuchi, General Chairman
• Professor Doo-Kwon Baik, Vice Chair
• Program Committee Members
• ISO/IEC JTC 1/SC 32/WG 2
• ISO TC 37
• ISO TC 184
• Thanks to Speakers
• Thanks to host and organizers

3
Thanks
Host IPSJ / ITSCJ Information Processing
Society of Japan /        Information Technology
Standards Commission of Japan. Supporter OGIS-RI
Co. Ltd. (Osaka Gas Information SystemsResearch
Institute) Sponsors Infoterm International
Information Centre for Terminology TermNet The
International Network for Terminology
Information System Society of Japan
UML based Modeling Technologies Promotion
4
Introduction to the Open Forum on Metadata
Registries 2006
1
K 1
Bruce Bargmeyer
Lawrence Berkeley National Laboratory University
of California Tel [phone],
bebargmeyer_at_lbl.gov
5
WG 2 Metadata
TC 37 Terminology
6
We have come to join
Terminology Metadata
7
???

• Questions from friends, relatives, associates TC
  184, TC 154, ebXML Asia (Reg/Rep), ODM, Dublin
  Core, practitioners,
• Now
• Who are WG 2/TC 37?
• How did WG 2/TC 37 earn a living?
• What skills and tools do they have?
• Future
• Can they work together?
• Can they earn a living in a changing world?
• What skills and tools do they need?
• What are the most promising directions?
• We will discuss these and hear about new RD,
  standards development, implementations, and new
  ideas from the speakers at OFMR2006.

8
Who is WG 1?Metadata

• Area of Work To develop and maintain standards
  that facilitate specification and management of
  metadata. Use of these standards will enhance the
  understanding and sharing of data, information
  and processes to support, for example,
  interoperability, electronic commerce and
  component-based development. The scope shall
  include
• a framework for specifying and managing metadata
  
• specification and management of data elements,
  structures and their associated semantics
• specification and management of value domains,
  such as classification and code schemes
• specification and management of data about
  processes and behaviour
• facilities to manage metadata, for example data
  dictionaries, repositories, information resource
  dictionary systems, registries and glossaries
• facilities to exchange metadata, including its
  semantics, over the Internet, intranets and other
  media.

9
Who is TC 37?Terminology and other language and
content resources

• Area of Work Standardization of principles,
  methods and applications relating to terminology
  and other language and content resources in the
  contexts of multilingual communication and
  cultural diversity.

10
What has WG 2 Focused On in the Past?

• Specification of the meaning of data (what data
  is meant to represent).
• Documentation of the provenance of data.
• Standardization and harmonization of data
• Stewardship of data

11
What has TC 37 Focused Onin the Past?

• Provide a systemic description of the concepts in
  the field of terminology
• Clarify the use of the terms in this field
• Addressed to, not only standardizers and
  terminologists, but to anyone involved in
  terminology work, as well as to the users of
  terminologies.
• Exchange of Terminology

12
Some Inspirational ISO TC 37 Standards

• ISO 704 Terminology work -- Principles and
  methods
• ISO 860 Terminology work -- Harmonization of
  concepts and terms
• ISO 1087 Terminology work -- Vocabulary -- Part
  1 Theory and application
• ISO 1087 Terminology work -- Vocabulary -- Part
  2 Computer applications
• ISO 12200 Computer applications in terminology --
  Machine-readable terminology interchange format
  (MARTIF) -- Negotiated interchange
• ISO 12620 Computer applications in terminology --
  Data categories
• ISO 16642 Computer applications in terminology --
  Terminological markup framework

13
How Did WG 2 People Earn a Living in the Past?

• Panhandle
• Public assistance (government leadership)
• Data Standards
• Data management
• Data administration
• Build and operate MDR
• Serve as MDR Registration Authority

14
How Did TC 37 People Earn a Living in the Past?

• Crumbs off library tables
• Catalog library contents
• Developing terminologies for application areas
• Computational linguistics

15
What Tools Skills did WG 2Use to Earn a Living?

• 11179 E 1
• Write data descriptions down in Text
• Used typewriters, word processors, or editors
• Started using DBMSs
• 11179 E 2
• Record data descriptions in databases
• SQL based query facilities
• Designed the Metadata Registry schema with
  modeling tools (but wrote it in text also)
• Used 11179 E 1 and E 2 Data Standards techniques

16
What Tools Skills did TC 37Use to Earn a
Living?

• Information science
• Write thesauri and taxonomies down in Text
• Used word processors, editors, or typewriters
• Started using DBMSs
• Software databaseData Category Registry

17
The Future?

• Data integration and harmonization is still a
  large challenge, but not exciting to large
  organizations.
• People like to make up new data and new words
  unfettered by the past. (It is fine for
  dictionaries and registries to record what they
  have done.)
• Metadata, thesauri and taxonomies are so last
  year.
• Now knowledge, semantics and ontologies are hot.
  They get increasing organizational mind share and
  funding.
• Knowledge bases
• Ontologies
• Triples (subject, verb, object)
• Inferencing
• Semantic Web
• There is recognition of the need for integration
  and harmonization.
• But, If you cant dance it, you cant teach it.
  from the movie Ballroom Dancing

18
New Tools

• Inference engines
• Reasoners
• Agents
• Triple stores
• Search engines
• Have these rendered metadata registries and Data
  Category Registries obsolete?
• Have these rendered TC 37 and WG 2 skills,
  techniques and technologies obsolete?

19
Semantics Where have we been?
Where are we planning to go?
System manuals
Semantic grids
Data dictionaries
Semantics services (SSOA)
11179 E1
Data ontology lifecycle management
Data Standards
11179 E2
Complex semantics management
Data Management/ Data Administration
ISO/IEC 11179 E3 19763 P 1-4 24707
Data engineering
Terminologies
Metadata Registries (MDR)
Semantic Web Ontologies
XML related standards
20
WG 2 Metadata
TC 37 Terminology
21
Ogdens Semiotic Triangle
Thought or Reference
Symbolises
Refers to
Symbol
Referent
Stands for
C.K Ogden and I. A. Richards. The Meaning of
Meaning.
22
Concept in Semiotic Triangle
Thought or Reference (Concept)
Symbolises
Refers to
Symbol
Referent
Stands for
Rose, ClipArt
C.K Ogden and I. A. Richards. The Meaning f
Meaning.
23
WG 2 Concept

• From 11179 E2 (2003)
• Concept unit of knowledge created by a unique
  combination of characteristics ISO 1087-12000,
  3.2.1
• Designation representation of a concept by a
  sign which denotes it ISO 1087-12000, 3.4.1
• Definition representation of a concept by a
  descriptive statement which serves to
  differentiate it from related concepts
  ISO 1087-12000, 3.3.1
• Concept Relationship a semantic link among two
  or more Concepts
• concept relationship type description a
  description of the type of relationship among two
  or more Concepts

24
WG 2 Concept
Figure 8  Data Element Concept metamodel
region ISO/IEC 11179 E2 (2003)
25
WG 2 Classification Scheme Item
Figure 7  Classification metamodel
region ISO/IEC 11179 E2 (2003)
26
TC 37 Concept
27
ConceptEssence and Differentia
TC 37 Definition
Ogden Symbol TC 37 Designation
(Sign)
28
ConceptEssence and Differentia
Definition Essence
Differentia
Ogden Symbol TC 37 Sign?
29
Rose

• 1. any of the wild or cultivated, usually
  prickly-stemmed, pinnate-leaved, showy-flowered
  shrubs of the genus Rosa. Cf. rose family.
• 2. any of various related or similar plants.
• 3. the flower of any such shrub, of a red, pink,
  white, or yellow color.
• --Random House Websters Unabridged Dictionary
  (2003)

30
Is each a Rose? as Defined by Essence and
Differentia
31
Concept Described by Relationships to Other
Concepts
Love Romance Marriage
CONCEPT
Refers To
Symbolizes
Rose, ClipArt
Stands For
Referent
32
SNOMED Terms Defined by Relationships

• Is this
• The thing that is defined as a procedure that
  involves an excision of a structure of lobe of
  lung? (Axiom)

2. A statement saying All procedures that
involve an excision of the structure of lobe of
lung are pulmonary lobectomy? (Falsifiable
proposition)
33
Rose Same Concept?
Romance Love Marriage XXX Baby Family
Romance Love Marriage XXX Baby Family
34
Rose Same Concept?
Romance Love Marriage XXX Baby Family
XXX Romance Love Marriage Baby Family
35
Rose Same Concept?
XXX
Romance Love Marriage XXX Baby Family
36
The Communication Process
CONCEPT
CONCEPT
Symbolises
Refers To
Refers To
Symbolises
Rose, ClipArt
Rose, ClipArt
Stands For
Stands For
Referent
Symbol
Symbol
37
CommunicationConcept vs. Symbol
Symbol
Symbol
CONCEPT
CONCEPT
Symbolises
Refers To
Refers To
Symbolises
I see a ClipArt image of a rose
Rose, ClipArt
Rose, ClipArt
Stands For
Stands For
Referent
Symbol
Symbol
Rose
Rose
38
RDF Symbol and Reference
Symbol
Symbol
CONCEPT
CONCEPT
Symbolises
Refers To
Refers To
Symbolises
I see a ClipArt image of a rose
Rose, ClipArt
Rose, ClipArt
Stands For
Stands For
Referent
Symbol
Symbol
39
RDF Both Symbols and Reference (Definition)
Edge
Node
Node
Subject
Predicate
Object
URI ..
Rose
URI ..
URI ..
40
Registry may be used to ground the Semantics of
an RDF Statement.
The address state code is AB. This can be
expressed as a directed Graph e.g., an RDF
statement
41
Grounding RDF nodes and relations URIs
Reference a Metadata Registry
dbAe0139
ai MailingAddress
dbAma344
ai StateUSPSCode
ABaiStateCode
_at_prefix dbA http/www.epa.gov/databaseA _at_prefix
ai http//www.epa/gov/edr/sw/AdministeredItem
42
URI Resolution in a Metadata Registry
Node and relationship meaning is established
through a URI pointing to an ISO/IEC 11179
Metadata Registry
Mailing Address
http//www.epa/gov/edr/sw/AdministeredItemMailin
gAddress

• The exact address where a mail piece is intended
  to be delivered, including urban-style address,
  rural route, and PO Box

State USPS Code
http//www.epa/gov/edr/sw/AdministeredItemStateU
SPSCode

• The U.S. Postal Service (USPS) abbreviation that
  represents a state or state equivalent for the
  U.S. or Canada

Mailing Address State Name
http//www.epa/gov/edr/sw/AdministeredItemStateN
ame

• The name of the state where mail is delivered

Needed Persistent URIs pointing to each item in
a 11179 Metadata Registry (Not currently part of
the standard).
43
Major Issues in Semantics Management addressed by
ISO/IEC 19763, 24707 and 11179

• Independent development and autonomous evolution
• Multiple ways to specify the same thing within a
  language (formalism, notation) and between
  languages
• Precise specification so that software (agents,
  applications, systems) can process without human
  intervention
• Harmonization and vetting within a community of
  interest
• Life cycle management (data, concept systems,
  ....)
• Processing based on semantic reasoning, rather
  than procedure

44
Strong Commonality of Purpose19763 24707
11179

• Semantics management - creating, managing,
  harmonizing, using, exchanging,
• Data,
• Concepts relationships (concept systems),
• Sentences/axioms,
• Created by diverse organizations,
• For diverse purposes
• Management approach coordinate and cultivate,
  rather than top-down command and control

45
ISO/IEC 19763Framework for Metamodel
Interoperability

• Objective
• Promote interoperability based on ontologies.
• Obstacles to ontology-based interoperation
• Issue 1
• Each ontology is developed independently and
  evolves autonomously.
• Issue 2
• Ontologies are described in several languages,
  sometimes with different names for the same thing
  in a Universe of Discourse or with the same name
  for different things in a UoD.
• FMI is to solve these problems, providing a
  registration framework for ontologies.

46
Difficulty caused by independent development
and autonomous evolution
This ontology has a definition of green card
and does not have a definition of Christmas
card.
This ontology does not have a definition of
green card but has a definition of Christmas
card.

• To avoid this difficulty, FMI Ontology
  Registration provides two types of ontologies,
  Reference Ontology and Local Ontology.

47
Reference Ontology

• FMI Ontology Registration provides the
  registration framework where a local ontology is
  defined based on reference ontologies

48
Goal of Common Logic

• Two agents, A and B, each have a first-order
  formalization of some knowledge
• A and B wish to communicate their knowledge to
  each other so as to draw some conclusions.
• Any inferences which B draws from A's input
  should also be derivable by A using basic logical
  principles, and vice versa
• The goal of Common Logic is to provide a logic
  based framework which can support this kind of
  use and communication without requiring complex
  negotiations between the agents.

49
Motivation for ISO/IEC 11179Metadata Registry
Extensions

• Support traditional data management and data
  administration in more powerful way.
• Go beyond traditional Data Standards and Data
  Administration. We want to support computer
  processing based on semantics--concepts and
  relationships.

50
Evolution of metadata technology

• From unstructured natural language metadata
  (written as text) to structured metadata
• Explicit modeling and characterization of
  relationships
• Graph based metamodels to aid comprehension and
  searching
• Formal ontologies
• AND from human consumption to machine processing
  for
• Software agents
• Computing inferences
• Semantic applications (e.g., transitive search,
  subsumption testing, etc.),
• Semantic services, E.g., mapping between
  equivalent value domains, units conversion,
• With new key technologies
• Graph databases (e.g., RDF) facilitate
  visualization machine processing
• Description logic (e.g., OWL DL) for more precise
  semantics machine reasoning
• Software Reasoners (e.g., inference engines)

51
ISO/IEC 11179Metadata Registry Extensions

• Register (and manage) any semantic artifacts that
  are useful for managing data.
• E.g., this includes registering concepts in any
  way related to data e.g., permissible values
  and data element definitions.
• It extends to registration of the full concept
  systems related to an organizations information
  held in structured, semi-structured or
  unstructured (text) form.
• E.g., may want to register keywords, thesauri,
  taxonomies, ontologies, axiomatized ontologies.
• Provide new services for semantic computing
  Semantics Service Oriented Architecture, Semantic
  Grids, semantics based workflows, Semantic Web .

52
ISO/IEC 11179Metadata Registry Extensions

• In addition to natural language, we want to
  capture semantics with more formal techniques
• First Order Logic, Description Logic, Common
  Logic, OWL
• However, maintain backward compatibility for
  implementers of 11179 E2

53
Motivation Urgent demands for Data Integration
and Harmonization

• Facilitate consolidation reorganization of
  government, private companies, and other
  organizations
• Ongoing acquisitions and mergers of organizations
• Corporations E.g, telecon, energy, banking,
• Government E.g., many agencies put under Dept of
  Homeland Security
• In National Institutes of Heath, the National
  Cancer Institute was created to focus on cancer
• Enable cooperation between countries and groups
• World Trade Organization
• North American Free Trade agreement
• European environment Basel Convention
• UN Food and Agriculture global food supply
• Enable sharing of data required quickly for
  emergencies
• Bird flu terrorism

54
Who could use extended metadata registries for
what purposes?

• Analysts, researchers anyone trying to create,
  harmonize, and manage data, concept systems,
  knowledge bases, rule bases, ontologies, RDF
  statements
• Engineering and Harmonization
• Vetting (gaining approval), establish trust, and
  enable stewardship
• Creators of new semantic computing systems
  applications
• Ground OWL ontologies and RDF statements
  (subjects, predicates, objects) in agreed upon
  definitions maintained in a metadata registry
• Use managed semantics within a community of
  interest
• Integrate existing semantics in new ways
• Improve semantics re-use
• Computers that are processing semantic computing
  applications
• Agents to access, map, and reason over data and
  concepts
• Applications that interact with both concepts in
  concept systems and data in databases.
• Grid computing - grid software can use the MDR
  XML representations for exchanging comparing
  objects (also, possibly RDF or OWL
  representations). Service Metadata in an MDR
  can be used on the grid to support semantic
  service discovery, service consolidation and
  dynamic creation of services workflows.

55
Concept Management

• In general, we want to register any concept based
  graph structure comprised of nodes,
  relationships, and possibly axioms
• possibly including millions of concepts, millions
  of terms, and millions of relationships (maybe
  billions).
• We want to link the concepts (e.g., research
  organization w, person x, disease y, location z)
  to data and text.

56
Example Concept Systems

• NBII Biocomplexity Thesaurus
• National Cancer Institute Metathesaurus
• NCI Data Elements (National Cancer Institute
  Data Standards Registry
• UMLS (non-proprietary portions)
• GEMET (General Multilingual Environmental
  Thesaurus)
• EDR Data Elements (Environmental Data Registry)
• USGS Geographic Names Information System (GNIS)
  HL7 Terminology, Data Elements
• Mouse Anatomy
• GO (Gene Ontology)
• EPA Web Registry Controlled Vocabulary
• BioPAX Ontology
• NASA SWEET Ontologies

57
Concept Systems and traditional metadata can be
represented queried as graphs
Nodes represent concepts or types of metadata
A
Lines (arcs) represent relationships
2
1
b
a
c
d
58
Finding Hidden Information in Registry Metadata
(Including Concept Systems)
Waterfowl
Waterfowl
Goose
Duck
Goose
Duck
59
Include Concept System Semantics in Metadata
Registries
Represent concepts and relationships as nodes
and edges in formal graph structures e.g., is-a
hierarchies.
Waterfowl
Duck
Goose
60
What new search capabilities do graph models
inference support?

• SQL-like structured queries (e.g., RDQL)
• e.g., SELECT ?x WHERE (?x rdftype
  xmdrValueDomain)
• Can span items that are only indirectly connected
• e.g., data elements associated with a permissible
  value
• Expand queries to subsumed classes in a hierarchy
• e.g., all cities within state and states within
  countries (partonomy)
• e.g., all species subsumed under birds
  (taxonomy)
• Search for higher level concepts or metadata
• e.g., all superclasses (ancestors) of a
  particular class
• e.g., least common ancestor (subsuming concept)
  for cat and snake
• Explore sibling items
• e.g., other airport codes comparable to SFO

61
Inference
Disease
is-a
is-a
Infectious Disease
Chronic Disease
is-a
is-a
is-a
is-a
Heart disease
Polio
Smallpox
Diabetes
Signifies inferred is-a relationship
62
Taxonomies partonomies can be used to support
inference queries
E.g., if a database contains information on
events by city, we could query that database for
events that happened in a particular county or
state, even though the event data does not
contain explicit state or county codes.
63
Relationship metadata can be used to infer
non-explicit data
Analgesic Agent

• For example
• patient data on drugs currently being taken
  contains brand names (e.g. Tylenol, Anacin-3,
  Datril,)
• (2) thesaurus connects different drug types and
  names with one another (via is-a, part-of, etc.
  relationships)
• (3) so patient data can be linked and searched
  by inferred terms like acetominophen and
  analgesic as well as trade names explicitly
  stored as text strings in the database

Non-Narcotic Analgesic
Analgesic and Antipyretic
Acetominophen
Nonsteroidal Antiinflammatory Drug
Datril
Anacin-3
Tylenol
64
Least Common Ancestor Query
What is the least common ancestor concept in NCI
Thesaurus for Acetominophen and Morphine
Sulfate? (answer Analgesic Agent)
Analgesic and Antipyretic
65
Example sibling queries concepts that share
ancestor

• Environmental
• "siblings" of Wetland (in SWEET ontology)
• Health
• Siblings of ERK1 finds all 700 other kinase
  enzymes
• Siblings of Novastatin finds all other statins
• 11179 Metadata
• Sibling values in an enumerated value domain

66
More complex sibling queries concepts with
multiple ancestors

• Health
• Find all the siblings of Breast Neoplasm
• Environmental
• Find all chemicals that are a
• carcinogen (cause cancer) and
• toxin (are poisonous) and
• terratogen (cause birth defects)

site neoplasms
breast disorders
Breast neoplasm
Non-Neoplastic Breast Disorder
Eye neoplasm
Respiratory System neoplasm
67
Metadata relationships can also be used to infer
connected information
Database

• For example
• An agency has hundreds of different databases,
  with metadata for each in a 11179 Registry
  .
• Manager asks which databases can be searched to
  find specific information for China?
• Search code values for China ( synonyms like
  CN) and show all databases that are connected
  only indirectly via Ennumerated Value Domain,
  Value Domain and Permissible Value

Data Element
Value Domain
Ennumerated Value Domain
Non-ennumerated Value Domain
Permissible Value
Permissible Value Meaning
68
Different ontologies support semantics management
at different levels
11179 classes, properties, and relationships
11179 Metamodel Level
Concepts and Terms
11179 Registry Level
Database B
Application Software Level
Database A
69
Nodes and relations support inference on 11179
metamodel
70
Advanced Use Scenario Allergy AlertLinking
concept searches, metadata searches, and database
queries (outline)

• Event Doctor prescribes medicine. Will patient
  have allergic reaction?
• Event triggers concept system search to determine
  if the prescription is a drug and if so, what
  type of drug. The first search is for an isA
  relation, followed by a search for a partonomy
  relation
• Then system must perform a metadata search to
  find data elements in information systems
  relating to patient allergy
• Result of metadata search enables a database
  lookup in a patient record
• Database lookup produces a drug reaction code
• System must look up the code in a concept
  systemto find type of reaction and category of
  drug
• Relate drug reaction to category of prescribed
  drug
• Produce Allergy Alert for Dr. Patient

71
Scenario Allergy Alert

• Event
• Prescription 500 mg Prevpac bid

72
Scenario Allergy Alert

• Is this a prescription for a drug?
• Yes concept system lookup says prescription
  category is for drugs and devices, That is,
  Prevpac isA Drug
• If so, what category (ies) of drug?
• Lookup in Concept system (partonomy) shows that
  Prevpac contains
• Lansoprozole - proton pump inhibitor
• Amoxicillin - beta-lactam antibiotic
• Clarithromycin - macrolide antibiotic

73
Scenario Allergy Alert

• Does the patient have an allergy to any of the
  drugs?
• Need to Metadata lookup to find relevant data
  elements in patient record databases
• Need to join the contents of the database(s)
• Diagnosis Allergy to ___________
• Observation Apparent reaction to ______

74
Scenario Allergy Alert

• Search Database Patient Record
• Result Dx ICD-9-CM code 996.2 Unspecified
  adverse effect of drug, medicinal and biologic
  substances
• Search Concept System
• Result Adverse reaction SNOMED
• 294461000 (antibacterial drug allergy)
• 246075003 (causative agent)
• 392284008 (nafcillin)

75
Scenario Allergy Alert
All of these contain a form of penicillin
76
Scenario Allergy Alert
Nafcillin isA Penicillin Amoxicillin isA
Penicillin
77
Alert

• Warning!!! Patient has had a prior adverse
  reaction to Nafcillin which is similar to the
  component Amoxicillin in the current
  prescription.
• Note The Rand Corporation states that billions
  of dollars per year can be saved in healthcare
  expenditures and better results can be achieved
  with improved medical systems of this type.
• --Rand Review, Fall 2005

78
Summary MDR FMI Concept System Store
Concept systems Keywords Controlled
Vocabularies Thesauri Taxonomies Ontologies Axioma
tized Ontologies (Essentially graphs
node-relation-node axioms)

79
Summary MDR FMI to Manage Concept Systems

Concept system Registration Harmonization
Standardization Acceptance (vetting) Mapping
(correspondences)
80
Summary MDR FMI for Life Cycle Management

Life cycle management Data and Concept
systems (ontologies)
81
Summary MDR for Grounding Semantics
Metadata Registries
Semantic Web RDF Triples Subject (node URI) Verb
(relation URI) Object (node URI)
Ontologies
82
11179 E3 Proposal
83
TC 37 WG 2Lets continue to align our standards

• Some topics to discuss
• What is understood under "concept" in
  terminology, ontology, metadata approach?
• Should WG 2 use the term concept system where
  we are using it?
• How do we include CL axioms and sentences?
• What is the role of attributes, characteristics,
  qualifiers, identifiers etc. in data modelling?
• What should be the use of "meta" metadata,
  metainformation, metaontology, metasystem, etc.
• What is the role of classification (incl.
  different types of classification)?
• Can we agree on harmonized naming principles?
• What is a typology, nomenclature, categorization?

84
Presentations to Follow

• I look forward to hearing the tutorials and
  presentations covering standards, RD, and
  practical applications in these areas.