20130905_feng_chia_gis_center_geospatial_ontology
TRANSCRIPT
1
Geo-Ontology and -Semantics: From Theoretics to Practices
Institute of Information Science, Academia Sinica
Faculty of Geo-Information Science and Earth Observation (ITC), University of Twente
Dongpo Deng
Tuesday, September 10, 2013
What is ‘O’ntology?
• The term originated from a philosophical discipline
• A branch of philosophy that deals with the nature and the organization of reality
• It was defined by Aristotle in Metaphysics, IV, 1
• It tries to answer the questions:
• What is being?
• What are the features common to all beings?
• How should things be classified?
2Tuesday, September 10, 2013
What is ‘o’ntology?
• a specification of a conceptualization (Gruber, T. 1993)
• a formal specification of a shared conceptualization (Borst et al., 1997)
• is a hierarchically structured set of terms for describing a domain that can be used as a skeletal foundation for knowledge base (Swartout et al., 1996)
• the method to extract a catalogue of things or entities (C) that exist in a domain (D) from the perspective of a person who use a certain language (L) to describe it (Sowa, 2000)
3Tuesday, September 10, 2013
What is ‘o’ntology?
• From AI perspective (Agarwal, 2005)
• “conceptualization” is explained as an abstract model of some phenomenon in the world by having identified the relevant concepts of that phenomenon.
• “explicit” means that type of concepts used and the constraints on their use are explicitly defined
• “formal” refers to the fact that the ontology should be machine-readable
• “shared” refers to notion that on ontology captures consensual knowledge
4Tuesday, September 10, 2013
An example of ontology description
• Example Vocabulary and meaning (“definitions”)
• A ‘Carnivore’ is a concept whose members are exactly those animals who eat only meat
• A ‘Bear’ is a concept whose members are a kind of ‘Carnivore’
• A ‘Cub’ is a concept whose members are exactly those ‘Bear’ whose age is less than one year
• A Panda is a individual of a ‘Bear’
5Tuesday, September 10, 2013
An example of ontology description
• Background knowledge/constraints on the domain (“general axioms”)
• No individuals can be both a Herbivore and a ‘Carnivore’
• Each ‘Bear’ has a period of ‘Cub’
• The age of adult ‘Bears’ is at least 1 year old
6Tuesday, September 10, 2013
The meaning triangle
7Tuesday, September 10, 2013
Syntax-Semantics-Pragmatics
• Syntax (語法) deals with the study of relationships between symbols
• Semantics (語意) analyzes the relationships between symbols and things in the real world they denote (referent)
• Pragmatics (語用) goes beyond syntax and semantics, and researches how symbols are used for particular purposes. Thus, is analyzes relationships between symbols and specific agents
8Tuesday, September 10, 2013
Ontologies for communication
Adapted by Maedche, 2001
9Tuesday, September 10, 2013
Ontology in Computer Science
• An ontology refers to an engineering artifact consisting of:
• A vocabulary used to describe (a particular view of) some domain
• An explicit specification of the intended meaning of the vocabulary
• almost always includes how concepts should be classified
• Constraints capturing additional knowledge about the domain
• Ideally, an ontology should:
• Capture a shared understanding of a domain of interest
• Provide a formal and machine manipulable model of the domain
10Tuesday, September 10, 2013
Why develop ontology?
• To share common understanding of the structure of information among people or software agents
• To enable reuse of domain knowledge
• To make domain assumptions explicit
• To separate domain knowledge from the operational knowledge
• To analyze domain knowledge
11Tuesday, September 10, 2013
Ontology spectrum
Logical Theory
Conceptual Model
Thesaurus
Taxonomy
is Disjoint Subclass of with Transitivity property
is Subclass of
has Narrower meaning then
is Sub-Classification of
Model LogicFirst Order Logic
Description LogicDAML_OIL, OWL
UML
RDF/SXTM
Extended ER
ERDB Schemas, XML Schema
Relational Model, XML
strong semantics
weak semantics
Semantic Interoperability
Structure Interoperability
Semantic Interoperability
Syntactic Interoperability
12Tuesday, September 10, 2013
物種分類名錄 (taibif.tw)
13Tuesday, September 10, 2013
OpenStreetMap Map Features
14Tuesday, September 10, 2013
ADL Gazetteer Feature typesFrom http://www.alexandria.ucsb.edu/~lhill/FeatureTypes/ver070302/00000250.htm
15
ADL Gazetteer Feature types
Tuesday, September 10, 2013
From http://wordnetweb.princeton.edu/ 16Tuesday, September 10, 2013
17
The application ontology inheriting GeoSPARQL
Tuesday, September 10, 2013
Edited by Protege
18Tuesday, September 10, 2013
19Tuesday, September 10, 2013
Different kinds of ontologies and their relationships
top-level ontologyvery general concepts, e.g. toopology, mereology, geometry, ...
vocabularies related to a generic domain by specializing the top-level ontologies, e.g. GeoSPARQL
domain ontology task ontology
application ontology
vocabularies related to a generic task or activity by specializing the top-level ontologies, e.g. Semantic Sensor Network (SSN) ontology
concepts inheriting domain or task ontologies for supporting in certain activities, e.g. we use SSN and GeoSPARQL to create an ontology for ecological observation
20Tuesday, September 10, 2013
Courtesy: Pascal Hitzler, GeoSemantics, 2009
Description Logics
Tuesday, September 10, 2013
Courtesy: Pascal Hitzler, GeoSemantics, 2009
Ontology and reasoning
Tuesday, September 10, 2013
Courtesy: Pascal Hitzler, GeoSemantics, 2009
Ontology and reasoning
Tuesday, September 10, 2013
Courtesy: Pascal Hitzler, GeoSemantics, 2009
Ontology and reasoning
Tuesday, September 10, 2013
Courtesy: Pascal Hitzler, GeoSemantics, 2009
Ontology and reasoning
Tuesday, September 10, 2013
Courtesy: Pascal Hitzler, GeoSemantics, 2009
Ontology and reasoning
Tuesday, September 10, 2013
Courtesy: Pascal Hitzler, GeoSemantics, 2009
Ontology and reasoning
Tuesday, September 10, 2013
What’s the (Geo) Problem?
• What’s special about Spatial?
• spatial-time-attributes
• What is geospatial interoperability?
• GML? WFS? or more alternatives?
• semantic Web - microformat tagging and (multiple) identity
• Semantic Web - (actionable) relationships and triple identity
• geosemantic - geotagging position
• Geosemantic - spatial(-temporal) theories, relationships, mediations, transformations
28Adapted from J. Lieberman (2007) Geospatial Semantic Web: Is there life after geo:lat and geo:long ?
Tuesday, September 10, 2013
What’s the (Geo) Problem?
• Feature (type) and Geometry (representation)
• Model dependencies
• Community of discourse
• Scale
• Reference frame / coordinate system
• Perspective
• Geospatial plus other (semantic) dimensions
29Tuesday, September 10, 2013
Two approaches for studying on geo ontologies
• There are two distinct approaches that applied ontology in GIScience.
• The philosophical approaches aim to identify specifications of top-level categories from a formal ontology perspective,
• The domain-specific and task-oriented approaches focus on explicating the actions, terms and relation for particular specification and ranging from natural language to rigorously formal specifications.
30Tuesday, September 10, 2013
How difficult to develop geospatial ontology?
• Geographic objects are typically complex, and they will in every case have parts. (Simons, 1987; Smith and Mark, 1998)
• The geographic domain has specific issues regarding ontology primarily because of its unstructured characteristics
• A standard terminology is not prevalent within the GIScience domain and is dependent on the context of use and the user
• causes confusion in specification of universally accepted entities, concepts, rules, relation, and semantics as the basis of a consensual ontology.
31Tuesday, September 10, 2013
Geographical Entities are Indeterminate and Ambiguous Objects
Philosophically speaking: Where does themountain begin and the valley end?
How can we derive a commonsemantics which can refer accuratelyto these kinds of objects?
B. Smith and D. Mark, 2003. Do Mountains Exist? Towards an Ontology of Landforms. Environment and Planning B, 30(3), 411-42732
Tuesday, September 10, 2013
Geographical Entities are Indeterminate and Ambiguous Objects
Philosophically speaking: Where does themountain begin and the valley end?
How can we derive a commonsemantics which can refer accuratelyto these kinds of objects?
B. Smith and D. Mark, 2003. Do Mountains Exist? Towards an Ontology of Landforms. Environment and Planning B, 30(3), 411-42732
Tuesday, September 10, 2013
Geospatial objects: vagueness and ambiguity
• Geographic categorization and classification are scale- and size-dependent,
• regionalization in space and time is human-dependent, and location and structure of boundaries shape many geographical categories.
• The ways that space and time determine relations and property inheritance are not yet clear.
• Human dependence means that geographic categories and nomenclature can have different meanings in different application contexts.
• fiat (tennis court) and bona fide (shoreline or riverbanks)
• open (bay) and closed (lake)
33B. Smith and D. Mark, 2003. Do Mountains Exist? Towards an Ontology of Landforms. Environment and Planning B, 30(3), 411-427
Tuesday, September 10, 2013
Conceptual model in General Feature Model (GFM)
34Tuesday, September 10, 2013
W3C Basic Geo Vocabulary
35http://www.w3.org/2003/01/geo/Tuesday, September 10, 2013
Map4RDF
36
http://oeg-dev.dia.fi.upm.es/map4rdf/
Tuesday, September 10, 2013
Geonames.org
37http://www.geonames.org/maps/google_25.048_121.532.htmlTuesday, September 10, 2013
‘Taipei’ in Geonames
38http://sws.geonames.org/1668341/about.rdf
Tuesday, September 10, 2013
Geonames ontology
39
http://lov.okfn.org/dataset/lov/details/vocabulary_gn.html
Tuesday, September 10, 2013
Ordnance survey Ontologies
40
http://data.ordnancesurvey.co.uk/ontology/
Tuesday, September 10, 2013
Spatial Relations in OS ontologies
41http://data.ordnancesurvey.co.uk/ontology/spatialrelations/contains
Tuesday, September 10, 2013
Spatial Relations in OS ontologies
42http://data.ordnancesurvey.co.uk/ontology/spatialrelations/contains.ttl
Tuesday, September 10, 2013
GeoSPARQL
• The GeoSPARQL will be a new OGC standard, which provides three main components for encoding geographic information:
• (1) The definitions of vocabularies for representing features, geometries, and their relationships;
• (2) A set of domain-specific, spatial functions for use in SPARQL queries;
• (3) A set of query transformation rules
43Robert Battle, Dave Kolas, 2011. Enabling the Geospatial Semantic Web with Parliament and GeoSPARQL
Tuesday, September 10, 2013
Components of GeoSPARQL
• Vocabulary for Query Patterns
• Classes
• Spatial Object, Feature, Geometry
• Properties
• Topological relations
• Links between features and geometries
• Datatypes for geometry literals
• ogc:wktLiteral, ogc:gmlLiteral
• Query Functions
• Topological relations, distance, buffer, intersection, …
• Entailment Components
• RDFS entailment
• RIF rules to compute topological relations
44Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Vocabulary: Basic Classes and Relations
45Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Details of ogc:wktLiteral
46Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Details of ogc:gmlLiteral
47Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Topological Relations between ogc:SpatialObject
48Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
RCC8, Egenhofer & Simple Features
49Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Example Data
50Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Why Encode Geometry Data as a Literal?
51Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Why don’t GeoSPARQL support W3C Basic Geo?
• Too simple to meet our requirements
• Can’t use different datums and coordinate systems
• Limited number of geometry types
• W3C Basic Geo data can easily be converted to wktLiteral
52Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Query Functions
53Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Query Functions
54Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Query Functions
55Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Query Functions
56Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Topological Query Functions
57Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Topological Query Functions
58Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Example Query
59Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL RDFS Entailment Extension
60Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Simple Features Geometry Types
61Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Query Rewrite Extension
62Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
GeoSPARQL Query Rewrite Extension
63Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Query Rewrite Rules
• Used to compute Feature-Feature spatial relations basedon default geometries
• Specified as a collection of RIF rules
• Example: ogcr:sfEquals
64Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
Summary of Conformance Classes
• Parameters
• Serialization
• WKT
• GML
• Relation Family
• Simple Features
• RCC8
• Egenhofer
65
Determines geometry classes and geometry literal datatype
Determines topology properties and topology functions
Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.Tuesday, September 10, 2013
Why don’t you support W3C Basic Geo?
• Too simple to meet our requirements
• Can’t use different datums and coordinate systems
• Limited number of geometry types
• W3C Basic Geo data can easily be converted to wktLiteral
66Courtesy from M. Perry (2012)The GeoSPARQL OGC Standard, Terra Cognita.
Tuesday, September 10, 2013
BBN Parliament
67Tuesday, September 10, 2013
68Tuesday, September 10, 2013
An example query of GeoSPARQL(1)
• Within
69
SELECT DISTINCT ?POO ?POO_wktWHERE { ?POO a eoe:PointOfObservation; geo:hasGeometry ?POO_geo. ?POO_geo geo:asWKT ?POO_wkt.
FILTER (geof:sfWithin(?POO_wkt, "Polygon((121 22, 121 23, 122 23, 122 22, 121 22))"^^sf:wktLiteral))
}
Tuesday, September 10, 2013
Query results (1)
70Tuesday, September 10, 2013
An example query of GeoSPARQ(2)
• buffer and within
71
SELECT DISTINCT ?p_wkt?POO_wkt ?distanceWHERE {
?POO a eoe:PointOfObservation; geo:hasGeometry ?POO_geo. ?POO_geo geo:asWKT ?POO_wkt. geo:point_238918712815615_439043552803129 geo:asWKT ?p_wkt;
LET (?buff := geof:buffer(?p_wkt, 3000, units:metre)) . FILTER (geof:sfWithin(?POO_wkt, ?buff)) . LET (?distance := geof:distance(?POO_wkt, ?p_wkt, units:metre))}
Tuesday, September 10, 2013
Query results (2)
72Tuesday, September 10, 2013
Q&A
73Tuesday, September 10, 2013