Ontologies for Earth System Science: Experiences in the Development of a Community Knowledge Base

Rob RaskinNASA/Jet Propulsion Laboratory

Pasadena, CA

Feb 22, 2008

Data to Knowledge

Data Information Knowledge

Basic Elements Bytes Numbers Models FactsServices Save Visualize Infer Understand PredictStorage File Database GIS Ontology MindVolume High LowDensity Low High

Syntax Semantics

What is Knowledge? Facts, relations, meanings, contexts,

common sense Information with context Shared understanding of meaning Suitable for reasoning/inference Dynamic, expandable

Application:Intelligent Search for Data Consults knowledge base to find

alternative meanings Clustered by: synonyms, parent, children

Enables discovery of resources without exact keyword match

Semantic understanding is crucial Common search engines (Google) use these

capabilities only minimally, at present

Application: Intelligent Search for Data (more) Noesis ontology-aided search tool

http://noesis.itsc.uah.edu Provides access to:

Data Journal articles Web pages Experts (people)

Semantic Understanding is Semantic Understanding is Difficult!Difficult!

Variable t: temperatureVariable t: time

Sea surface temperature: measured 3 m above surface Sea surface temperature: measured at surface

Data quality= 5 Data quality= 3

Let’s eat, Grandma.Let’s eat Grandma.

Time flies like an arrow.Fruit flies like a pie.

Semantic Understanding is Semantic Understanding is Difficult!Difficult!(more)(more)

LA Times headline

“Mission accomplished. Major combat operations in Iraq have ended”

-Pres. Bush, 2003

“not a nobody” “ain’t nobody” “yeah, right”

Knowledge Base of Facts as Triples

Subject-Verb-Object representation

Flood is a WeatherPhenomena GeoTIFF is a FileFormat Soil Type is a PhysicalProperty Pacific Ocean is a Ocean

Ocean has substance Water Sensor measures Temperature

Semantic Web Vision Web page creators place XML tags around

technical terms on web pages XML tags point to knowledge base

(namespace) where term is “defined” This vision has not caught on, in part due to

the success of free-text searching without semantic aid

Ontologies

Method to store “facts” General definition: “all that is known” Computer science definition: Machine-readable definition

of terms and how they relate to one another As with a dictionary, terms are defined in terms of other terms

Provide shared understanding of concepts Support knowledge reuse Support machine-to-machine communications with

deeper semantics than controlled vocabulary

Ontology Properties

Standard languages make it easy to extend (specialize) concepts developed by others

Synonym support (multiple terms with same meaning) Label available to indicate preferred term for each

community Homonym support (multiple meanings of

same term) Separate namespaces (President:Bush vs Plant:Bush)

Plate tectonics - before

Plate Tectonics Ontology

Atmosphere Ontology…

Atmosphere Ontology

Ontology Languages:RDF and OWL

W3C has adopted XML-based standard ontology languages

Resource Description Formulation (RDF) Ontology Web Language (OWL)

Languages predefine specific tags RDF: Class, subclass, property, subproperty, … OWL: Extends RDF to predefine further tags such as cardinality

Three flavors of OWL (Lite, DL, and Full)

Use of standard languages make it easy to extend (specialize) work of others

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

Non-LivingSubstances

LivingSubstances

PhysicalProcesses

Earth Realm

PhysicalProperties

Time

NaturalPhenomena

Human Activities

Integrative Ontologies

Space

Data

Faceted Ontologies

Units

Numerics

SWEET 1.0 Ontologies

SWEET 2.0 Ontologies: Modular Design

Mathematics Units

Electric/Magnetism

TimeSpace

RadiationTransfer

GeophysFluid Dynam

HumanActivities

ClimateChange

Troposphere

Thermo

Land Surface

Waves

Heliosphere Cryosphere

Mechanics

BasicScience/Math

SupportingGeophysicalPhenomena

PlanetaryRealms

ApplicationsAir Pollution

WaterResources

Geosphere

Ecosphere

Biogeo-chemistry

PlanetaryStructures

import

OceanUpper

Atmosphere

Geo-magnetism

PlanetaryGravity

BiogeochemCycles

Energy etc.

Chemistry

SWEET as an Upper Level Earth Science Ontology

Math Physics Chemistry

Space

TimeProperty

EarthRealmProcess, Phenomena

SubstanceData

StratosphericChemistry

Biogeochemistry Specialized domains

import

import

SWEET

SWEET Numerical Ontologies

Intervals, numeric relations (<,>) Cartesian products Functions, derivatives Fuzzy concepts

“near” Spatial concepts

0-D, 1-D, 2-D, and 3-D objects Coordinate systems Above, inside, etc.

Temporal concepts Instant, durations, geological time scales

SWEET Data Ontology Dataset characteristics

Format, data model, dimensions, … Provenance

Source, processing history, … Parameters

Scale factors, offsets, … Data services

Subsetting, reprojection, … Quality measures Special values

Missing, land, sea, ice, ...

Web Services Ontology OWL-S

Enables semantic descriptions of Web service inputs and outputs

Provides binding to actual implementation of Web service

Best Practices Keep ontologies small, modular

Be careful that “Owl:Import” imports everything Use higher level ontologies where possible

Identify hierarchy of concept spaces Model schemas Try to keep dependencies unidirectional

Gain community buy-in Involve respected leaders

PlanetOnt.org

Tec

hn

olo

gy

Geospatial semantic services

established

Geospatial semantic services proliferate

Scientific semantic assisted services

SWEET 3.0 with semantic callable

interfaces via standard programming languages

SWEET core 2.0 based on best practices decided from community

Scientific reasoning

Reasoners able to utilize

SWEET 4.0

Local processing + data exchange

Basic data tailoring services (data as

service), verification/ validation

Interoperable geospatial services

(analysis as service), explanation

Common vocabulary based

product search and access

Inte

rop

erab

le

Info

rmat

ion

In

fras

tru

ctu

re

Ass

iste

d

Dis

cove

ry &

M

edia

tio

n

Metadata-driven data fusion

(semantic service chaining), trust

Semantic agent-based integration

Semantic agent-based

searches

Geospatial reasoning, OWL-

Time

Cap

ab

ilit

yR

esu

lts

Improved Information

Sharing

Increased Collaboration & Interdisciplinary

Science

Acceleration of Knowledge Production

Revolutionizing how science is

done

RDF, OWL, OWL-S

Semantic Web Roadmap

Current Near Term Mid Term Long Term

Autonomous inference of

science results

Numerical reasoning

Semantic geospatial search & inference, access

SWEET core 1.0 based on GCMD/CF

Vo

cab

ula

ryL

ang

uag

es/

Rea

son

ing

Ou

tpu

tO

utc

om

e

Ontology Development Tools: CMAP Free, downloadable tool for knowledge

representation and ontology development

Visual language with input/export to OWL Supports subset of OWL language

http://cmap.ihmc.us/coe

Other Ontology Editors Public Domain

SWOOP Mindswap Lab (UMBC) http://www.mindswap.org/2004/SWOOP

Protégé Stanford http://protege.stanford.edu

Commercial Cerebra Construct

http://cerebra.com TopBraid Composer

http://www.topquadrant.com/topbraid/composer

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