Docstoc

Ontology-Summit-An-Ontology-Framework-NIST,-Gaithersburg,-MD-

Document Sample
Ontology-Summit-An-Ontology-Framework-NIST,-Gaithersburg,-MD- Powered By Docstoc
					                Ontology Summit:
             An Ontology Framework

                  NIST, Gaithersburg, MD
                     April 22-23, 2007

                             Michael Gruninger
                        University of Toronto, Canada

                                Leo Obrst
                         MITRE, McLean, VA, USA

University of Toronto        November 9, 2011
Agenda

• A Brief Inclusive Characterization of Ontologies
• A Common Set of Characteristics for Ontologies
• Ontology Dimensions:
   – Semantic
      • Degree of Structure and Formality
      • Expressiveness of the Knowledge Representation Language
      • Representational granularity
   – Pragmatic
      •   Intended Use
      •   Role of Automated Reasoning
      •   Descriptive vs. Prescriptive
      •   Design Methodology
• Conclusions


                                                                  2
A Brief Inclusive Characterization of Ontologies
(in Computer Science & Engineering)

• Ontologies are used to support sharable and
  reusable representations of knowledge
  – An early definition of Ontology: “a specification of a
    conceptualization” (Gruber, 1994)
• Nevertheless, the sheer range of current work in
  ontologies:
  – Including taxonomies, thesauri, topic maps, conceptual models,
    and formal ontologies specified in various logical languages
  – Raises the possibility of ontologies being developed without a
    common understanding of their definition, implementation and
    applications
• Our objective:
  – To provide a framework that ensures that we can support diversity
    without divergence
  – So that we can maintain sharability and reusability among the
    different approaches to ontologies
                                                                        3
A Common Set of Characteristics for Ontologies


• An ontology includes:
  – A vocabulary together with a specification of
  – The meanings of the terms in the vocabulary
• This specification includes:
  – Identification of the fundamental categories in the
    domain
  – Identification of the ways in which members of the
    categories are related to each other
  – Constraining the ways in which the relationships can be
    used



                                                              4
Ontology Dimensions:
Semantic & Pragmatic
• We propose a set of dimensions that can be used to
  distinguish among different approaches
  – Semantic
  – Pragmatic
• Semantic Dimensions: These constrain how a given
  approach specifies the meaning of the terms
  1) Degree of Structure and Formality
  2) Expressiveness of the Knowledge Representation Language
  3) Representational granularity
• Pragmatic Dimensions: These cover the context in which
  the ontology is designed and used
  1) Intended Use
  2) Role of Automated Reasoning
  3) Descriptive vs. Prescriptive
  4) Design Methodology


                                                               5
Ontology Semantic Dimension:
(1) Degree of Formality (Structure)
• Related to but not the same as the expressive power
  required of a representation language used to specify the
  ontology
• Informal: An ontology can be specified in English or some
  other natural language in a document
   – This is an informal ontology, although it can be rich, unambiguous, precise
• More Formal: A taxonomy can be term or concept based
   – Term-based: A topic hierarchy from more general terms at the top of the
     hierarchy to more specific terms as one descends through the hierarchy;
   – Concept-based: A hierarchy of classes in which the necessary and
     distinguishing properties of classes and their subclasses are represented
   – But, even if formalized, these models are very simply structured, i.e.,
     structured with a subsumption relation (narrower_than or subclass)
• Very Formal: An ontology for engineering equations
   – Would specify the formal semantics of its terms (such as quantity and unit
     of measure) in a language enabling precision & unambiguous expression
• Degree of Structure: that by which the vocabulary in an
  ontology is constrained and can support computation                              6
Ontology Semantic Dimension: (2) Expressiveness
of the Knowledge Representation Language

• This is a dimension partially dependent on the first
   – Since an informal ontology will be expressed only as a list of terms or
     enumerated definitions in a natural language such as English
• Although not a characteristic in itself for a given content,
  the language or framework that the content is modeled in:
   – Greatly affects, and in fact
   – Is the primary constraint on what can be expressed by the content
• Predicates and individuals/instances, for example cannot
  be expressed in propositional logic
• Many of the notions termed “ontology” cannot be
  expressed formally
• Similarly, relationships besides subclass or narrower-than
  cannot be expressed in a strict or strong (i.e., at least
  partially formalized) taxonomy, although they can be in a
  less strict (informal) taxonomy
• In general, only if the KR language is logic-based will the
  ontology being machine-interpretable                                         7
Ontology Semantic Dimension:
(3) Representational Granularity:
• An ontology may contain terms and limited inter-
  relationship representation
   – Example: a simple taxonomy
   – Example: a very formal ontology expressed in Common Logic but which
     contains only 3 classes and 2 properties
• Or it may contain much more detail including many
  restrictions concerning how terms can relate to each other
   – Example: a very detailed English description of biological classes and
     discriminating properties
   – Example: a very formal ontology expressed in KIF, CL, Cyc-L, or
     OWL+SWRL which contains thousands of classes, thousands of
     properties, thousands of rules, and billions of instances/individuals
• Some quantifiable metrics may give indications of the
  representational granularity:
   – Examples: average subclass/subproperty depth, average
     density/bushiness, number of axioms, etc.


                                                                              8
Ontology Pragmatic Dimension:
(1) Intended Use, Application Focus
• The intended use may be:
   • To share knowledge bases
   •   To enable communication among software agents,
   •   To help integrate disparate data sets
   •   To represent a natural language vocabulary
   •   To help provide knowledge-enhanced search
   •   To provide a starting-point for building knowledge systems
   •   To provide a conceptual framework for indexing content, etc.
   •   The intended use often means that:
• Typically, there is some application that is envisioned for which the
  ontology is being developed
   • Categorization: one might want to situate documents within a framing topic
     taxonomy that roughly characterizes the primary content of the document and helps
     one semantically loosely organize document collections
   • Search Enhancement: one might want to use a thesaurus, its synonyms and
     narrower-than terms, to enhance a search engine that can employ query term
     expansion, expanding the user‟s text search terms to include synonyms or more
     specific terms and thus increasing the recall (i.e., total set of relevant items) of
     retrieved documents
   • Enterprise Modeling, Question-Answering, Semantic Web Service Discovery
   • Semantic Search: one might to use “concept” based rather than “term” based
     search in specific domains or across domains
   • Complex Decision-Making, Intelligence Analysis: requiring machine reasoning            9
Ontology Pragmatic Dimension:
(2) Role of Automated Reasoning
• Automated reasoning can range from simple to
  complex
• Simple automated reasoning can mean:
  • Machine semantic interpretability of the content, which only requires
    that the language that the content is modeled in is a logic
  • This is a principled or standards-based approach
• Or it can mean that a special interpreter/inference
  engine has been constructed that knows how to
  interpret the content
  • This is an ad hoc and often proprietary approach




                                                                            10
Ontology Pragmatic Dimension:
(2) Role of Automated Reasoning (cont‟d)
• Simple automated reasoning:
  • The machine may be able to make inferences using the subclass
    relation by which properties defined at the parent class are
    inherited down to the children classes; this is the property of
    transitivity
• More complex reasoning:
  • The machine may be able to take an arbitrary assertion in the KR
    language and classify it with respect to the taxonomic backbone of
    the ontology; e.g., description logics perform classificational
    reasoning
• Very complex automated reasoning:
  • The use of deductive rules, i.e., inference rules or expressions that
    combine information from across the ontology
    •   These characterize dependencies much like if-then-else statements in
        programming languages
    •   Business rules that try to characterize things that have to hold in an enterprise
        but which can‟t typically be expressed in relational databases or object models
    •   A logic-based KR language needed: „validly concludes‟ or „X is consistent with
        Y‟ are not expressible generally in ad hoc implementations
  • Theorem-proving, theorem-generation, etc.                                               11
Ontology Pragmatic Dimension:
(3) Descriptive vs. Prescriptive
• Descriptive:
  • Does the content describe, i.e., characterize the entities and their
    relationships as a user or an expert might characterize those objects?
  • Descriptive often takes a looser notion of characterization, perhaps
    allowing arbitrary objects into the model, which might not exist in the real
    world but which are significant conceptual items for the given user
    community
  • Potential partial synonym: Multiplicative, i.e., concepts can include
    anything that reality seems to require or any distinction that is useful to
    make
• Prescriptive:
  • Does the content prescribe, i.e., mandate the way that those entities and
    their relationships are characterized?
  • Prescriptive often takes a stricter notion of characterization, stating that
    only objects which actually exist or that represent natural kinds or types of
    things in the real world should be represented in the content of the
    engineering model
  • Potential partial synonym: Reductionist, i.e., concepts are reduced to the
    fewest primitives from which it is possible to generate complex reality

                                                                                    12
Ontology Pragmatic Dimension:
(4) Design Methodology
• Was there a methodology employed in the construction of the
  ontology? Possible ranges of methodology include:
• Bottom-up: A bottom-up (sometimes called “empirical”) methodology
  places strong emphasis on:
  • Either solely analyzing the data sources so that the resulting ontology
    covers their semantics
  • Or enabling arbitrary persons to characterize their content as they
    personally see fit, using terminology or metadata and whatever structuring
    relations (or not) that they desire to use, with perhaps an auxiliary notion or
    assumption that in by doing so, patterns of characterizations may emerge
    or be preferred by a large group or community of persons
  • Can be conceptually profligate, tolerable of redundancy and partial overlap
• Top-down: A top-down (sometimes called “rationalist”) methodology
  places strong emphasis on:
  • Developing the ontology using known notions about the world or domain
  • Independent of existing data sources whose semantics will be covered by
    the resulting ontology
  • Considering a range of questions that a domain expert might want to ask
    about the domain
  • Typically preferring a rigorous methodology focused on consistency,
    parsimony, etc.                                                                   13
Other Dimensions?

• Semantic Focus (originally: Concept-based?):
  – Term (0) – Concept (1) – Real World Referent (3)
  – Example: thesauri & some taxonomies: Term
  – Example: some conceptual models, some taxonomies, some
    ontologies: Concept
  – Example: some ontologies: RW Referent
• Precision: subsumed by Granularity?
• Scope: subsumed by composition of Application
  Foci?
• Human-coded vs. Machine-generated
• Range of values for the Ontology Types?
  – Scale?
  – Description only?
  – Both?
                                                             14
Conclusions

• We want to avoid positing definition, allow
  definition to emerge
• The framework is intended to offer dimensions for
  comparison
  – How our “ontologies” are alike, how they are different but
    comparable
• For too long we have been embroiled in endless
  argumentation
• Let‟s cooperate, find correspondences, figure out
  how we can semantically interoperate
• We‟ll have a number of presentations &
  discussions to help bring the communities
  together
• Goal: Converge!                                                15

				
DOCUMENT INFO
Shared By:
Tags: Ontol, ogy-S
Stats:
views:15
posted:12/1/2009
language:English
pages:15
Description: Ontology-Summit-An-Ontology-Framework-NIST,-Gaithersburg,-MD-