Transcript Document

Ontology Best Practices: Experiences with SWEET

Rob Raskin NASA/Jet Propulsion Laboratory Pasadena, CA

Why an Upper-Level Ontology for Earth System Science?

Why cooperate?

  Many common concepts used across Earth Science disciplines (e,g, Temperature, Pressure)    Provides common definitions for terms used in multiple disciplines or communities Provides common language in support of community and multidisciplinary activities Provides common “properties” (relations) for tool developers Reduced burden (and barrier to entry) on creators of specialized domain ontologies  Only need to create ontologies for incremental knowledge

Role of Upper Level Earth Science Ontology

Space Time Math Physics Chemistry

import

Property PlanetaryRealm Process, Phenomena Substance Data

import

Stratospheric Chemistry Biogeochemistry General domains

Common Earth elements

Specialized domains

Semantic Web for Earth and Environmental Terminology (SWEET)

     Concept space written in OWL Initial focus to assist search for data resources  Funded by NASA Later focus to serve as community standard Enables

scalable classification

of Earth system science concepts Populated initially with GCMD, CF concepts (decomposed)

SWEET 1.0 Ontologies (and their interrelationships) Living Substances Non-Living Substances

Faceted Ontologies Integrative Ontologies

Natural Phenomena Physical Processes Earth Realm Human Activities Data Physical Properties Space Time Numerics Units

SWEET 2.0

 Same facets, but organized by subject  12 ontologies --> 100 ontologies  Easier for domain specialists to build self-contained specialized ontologies that extend existing ones

SWEET 2.0

Ontologies

Importationt

Common Issues

     Units  UDUnits Standard math  Ordered pairs and triples, arithmetic operations Intervals  hasLowerBound, hasUpperBound, hasUnit Provenance  Sequence of steps Fuzzy concepts  nearlySameAs, similarityMeasure [0…1]

Best Practices (1):

   Identify characteristic level of abstraction of each term  If multiple definitions/levels (e.g., “climate”), repeat in multiple ontologies (namespaces) Keep ontologies small, modular  Be careful that “Owl:Import” imports everything  Use higher level ontologies where possible Identify hierarchy of concept spaces  Try to keep dependencies unidirectional

Best Practices (2):

 For synonyms, identify (community, preferred term) pairs  Gain community buy-in  Involve respected leaders  Most ontologies can be

faceted

Holistic

ontologies can be layers/wrappers atop faceted ontologies

Best Practices (3):

  Use OWL individuals (instances) sparingly   Assume OWL-DL will be used, because most tools cannot support OWL-Full Typically, a data collection is a “class” and a component of the Earth is a class  A particular observation at a specific time is a “state” (of the planet) which could be an individual OWL has limited capabilities   Instructions to reasoners can be included (e.g., “multiply”) Collect suggestions for implementations in future versions, or an OWL-Sci package

Community Issues

  Review Board  Who will oversee and maintain for perpetuity (or at least through the next funding cycle) ESSI? Content        Maintain alignment given expansion of classes and properties No removal of terms except for spelling or factual errors Subscription service to notify affected ontologies when changes made Must avoid contradictions Additions can create redundancy if sameAs not used Humans must oversee “matching” CF has established moderator to carry out analogous additions

PlanetOnt.org

Collaboration Web Site

       Discussion tools  Blog, wiki, moderated discussion board Version Control/ Configuration Management Trace dependencies on external ontologies Tools to search for existing concepts in registered ontologies Ontology Validation Procedure  W3C note is formal submission method Registry/discovery of ontologies Support workflows/services for ontology development

PlanetOnt.org

ESIP Federation

PO.DAAC Knowledge Bases

Public access Documents People Roles/Tasks Data Processing Data Products Science Concepts Metadata Missions Applications Announce ments Tools/ Services Instruments Inquiries Web Pages Organiza tions Computers

Resources

    SWEET  http://sweet.jpl.nasa.gov

Ontology development/sharing site  http://PlanetOnt.org

Noesis (search tool)  http://noesis.itsc.uah.edu

SESDI  http://sesdi.hao.ucar.edu