1 A partial ontology for soil investigation Overview - - PDF document

SLIDE 1

1

02.10.2003 | Folie 1 02.10.2003 | Folie 2

W G 14 2 ONTOLOGY

Herbert Schentz

I ncluding USE CASES from : I ntegrated Assessm ent Models & environm ental observation

Thom as Dirnböck

02.10.2003 | Folie 3

Overview

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off
• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

WG 142 Ontology

02.10.2003 | Folie 4

Definition

(http://en.wikipedia.org/wiki/Ontology_%28computer_science%29)

Part of the definition within Wikipedia … a formal logical ontology is specified as consisting of the following logical elements: concepts (classes, objects, or categories) … a formal logical ontology is specified as consisting of the following logical elements: concepts (classes, objects, or categories) with their characteristics (attributes, slots, functions, roles, or properties) and relations (generalization and specialization, functions) … a formal logical ontology is specified as consisting of the following logical elements: concepts (classes, objects, or categories) with their characteristics (attributes, slots, functions, roles, or properties) and relations (generalization and specialization, functions) restrained by logical axioms (assertions) and exemplified by instances of classes and specific properties…. … a formal logical ontology is specified as consisting of the following logical elements: concepts (classes, objects, or categories) with their characteristics (attributes, slots, functions, roles, or properties) Wikipedia

02.10.2003 | Folie 5

Overview

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

?mpf ?!

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

WG 142 Ontology

02.10.2003 | Folie 6

A sm all cat and m ice ontology mammal (maximalSize) cat mouse is a is a leg tail has 4 (maximum) has 1 (maximum) some hunt WG 142 Ontology

SLIDE 2

2

02.10.2003 | Folie 7

Administrative unit Administrative unit Administrative unit

A partial ontology for soil investigation

heavy metal ion pop site selection Site parameter soil sampling Sampling parameter transportation transport parameter Administrative unit Geological region Phases sorting parameters Codelist & Rules preparation Methodes ISO decomposition parameter Cadmium chromatographie wet analysis ......

Relations might be: Follows, partOf, produces, isIn, ....

Analysis specification

WG 142 Ontology

02.10.2003 | Folie 8

Overview

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

?mpf ?!

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

Aha ! WG 142 Ontology

02.10.2003 | Folie 9

Consequence You all have produced, worked with, needed, described, read, ... ontologies without knowing that those were ontologies.

So what is so exciting about

• ntologies ?

WG 142 Ontology

02.10.2003 | Folie 10

Form al languages

There are form al, human and machine readable

languages to notify ontologies:

• 1. The class m odel of UML (not so well m achine

readable)

• 2. RDF / RDFs (Resource description File – W3C

Standard)

• 3. OWL (Web Ontology language – W3C Standard)

We can exchange, We can exchange, share We can exchange, share and harmonize

• ntologies !

WG 142 Ontology

02.10.2003 | Folie 11

w hy Ontologies

Knowledge representation: The usual habitat of an eagle-owl is .... Expanded Thessauri: Controlled vocabulary + a lot

• f relations between the terms

Metadata: More detailed through the description of relations and restrictions Interface Descriptions: Extending the structural description of interfaces by descriptions of the contents

Ontologies: Unified formal descriptions of the meaning of the things, we want to share, exchange, use as inputs of models, ...

Semantic web WG 142 Ontology

02.10.2003 | Folie 12

Overview

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off
• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

WG 142 Ontology

SLIDE 3

3

02.10.2003 | Folie 13

Com m on Ontologies for sem antic Dataintegration

A Common Ontology is the key for data sharing and data integration

Distributed Data Distributed Applications (models) Workflow Engine Local Datamining View Data Portal

A Common Ontology WG 142 Ontology

02.10.2003 | Folie 14

W hy m odels?

Models combine knowledge from

various disciplines in an analytical framework

Assess the socio-econom ic and

environmental consequences of human activites

Derive „If-Then“ futures by

im plementing key processes – Scenarios

Inform decision makers about

consequences and options Integrated Assessment Models

02.10.2003 | Folie 15

I ntegrated Assessm ent ( I A) Models and Scenarios – som e exam ples

I A Models

• Environment-Energy Sector

RAINS (I I ASA SO2, NOx, NH3 Emission

und Effekte)

IPCC multi model approach

(I MAGE, AI M, ASF, MARI A, MESSAGE, MiniCAM, ..)

• Environment-Economy Sector

GTAP (Global Trade Analysis) WorldScan (e.g. policies for

implementing the Kyoto protocol)

• Sustainable development

WORLD 3 („Limits to growth“ and

„Beyond the limits“)

TARGETS (Tool to assess regional and

global environmental and health targets for sustainability)

Source: „Cloudy cristal balls“ report EEA

Scenarios

• ECN (1995) Energy scenarios for a

changing Europe

• CPB (1997) Economy and physical

environment)

• OECD (1997) The world in 2020: towards

a new global age

• IPCC (2000) SRES Emission Scenarios
• VISIONS 4. EU Framework Program
• Millenium Ecosystem Assessment (1998-

2005) Ecosystem services and human well being

ECN – Energy Research Centre of the Netherlands CPB – Dutch Central Planning Bureau

02.10.2003 | Folie 16

Advantages of I A and needs from environm ental observation

Advantages of IA

Reveals consequences of and options for policy Reveals interactions between different

environmental issues and the society

Rises public awareness of complex and long-

term, global environmental problems

Needs from environmental observation

Need for integration of different scales Need for integration of data Need for integration of knowledge Need for evaluation and validation

Integrated Assessment Models

02.10.2003 | Folie 17

I ntegration of data and m odels

Distributed ecological Data Distributed Applications Distributed Data mining With local tools

Species prediction model Calculation

• f

trajectories Groundwater flow model Massflow model Portal

Integrated Assessment Models

02.10.2003 | Folie 18

I ntegration of know ledge

The reliability of integrated assessments is based on the

use of multidisciplinary methods and the involvement of non-scientists

Participatory methods Scenarios IA models

• Mathematician
• System modeller
• Experts from social

and natural sciences

• etc.
• Stakeholders
• Politician
• Scientists
• Mediator
• etc.
• Social scientist
• Politician
• Lay experts
• NGO
• etc.

Integrated Assessment Models

SLIDE 4

4

02.10.2003 | Folie 19

Overview

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off
• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

Project 2006-2066 ???

WG 142 Ontology

02.10.2003 | Folie 20

How can w e get to a com m on

• ntology

It is im possible to establish one comm on ontology

for ecology in one step.

It is possible to establish a commonly agreed core

• ntology which can be kept stable.

It is possible to build domain ontologies (species

lists, meteorlogical ontology, air measurement

• ntology, ...) which are based on the core
• ntology.

Ontologies must always be extansible . (Science

never stops ! ... ) In .OWL e.g. this can be done a little bit m ore easily than in an XML Schema (XSD). WG 142 Ontology

02.10.2003 | Folie 21

Ontologies com m ing together

Establish local Ontologies Map between Ontologies Establish standards

SUMO (IEEE) DOLCE SICoP ?

Merge Ontologies

Commit to Ontologies

Bring in standards

ISO 19115 EML ABCD

WG 142 Ontology

02.10.2003 | Folie 22

Overview

• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off
• Definition
• Examples
• Why Ontologies?
• Common Ontology for semantic

Dataintegration

• Establishing Ontologies
• Invitation for the Kick Off

WG 142 Ontology

02.10.2003 | Folie 23

Ontologies are quite new

So the working group just has to be brought

together

and has to cooperate with ... Other WG like „data

model“

You are invited to come to the stand at the market

place. Or send an e-mail to: herbert.schentz@umweltbundesamt.at Thank you WG 142 Ontology