'Ontology Design Patterns' - User contributions [en]

WOP:2017

2017-07-13T07:13:48Z

RinkeHoekstra: /* WOP2017 Chairs */ Changed my affiliation

== 8th Workshop on Ontology Design and Patterns - WOP2017 ==

This is the eight edition in a series of workshops addressing the topic of '''ontology and semantic web patterns as best practices''', related to the [[Ontology_Design_Patterns_._org_(ODP) |''ontologydesignpatterns.org'']] initiative.

The Workshop on Ontology Design and Patterns targets topics relating to high quality ontology design. The workshop series addresses topics centered around quality in ontology design as well as ontology design patterns (ODP) in Semantic Web data and ontology engineering. ODPs have seen a sharp rise in attention in the past few years, both pertaining to this workshop series and other related events. Patterns can benefit knowledge engineers and Semantic Web developers with a direct link to requirements, reuse, guidance, and better communication. They need to be shared by a community in order to provide a common language, hence the aim of this workshop is twofold: 1) providing an arena for discussing patterns, pattern-based ontologies, systems, datasets, etc., and 2) broadening the pattern community by developing its own 'discourse' for discussing and describing relevant problems and their solutions. Related to the latter aim we see that it is an opportune time to open up the workshop to other approaches focusing on high quality ontology design, e.g. other methods and tools, with the intention to cross-fertilise these with the ODP idea.

For more background on the workshop series, see the [[WOP:Main | main page]].

WOP2017 is planned to be a ''full-day workshop'' consisting of three parts: paper presentations, pattern presentations and posters, and an interactive pattern writing and/or discussion session. The poster session will feature pattern posters, i.e. presentations of patterns submitted according the [[WOP:2017#Call_for_Patterns| call for patterns]].

__TOC__

== Venue and dates ==
The eight edition of WOP will be held in conjunction with [http://iswc2017.semanticweb.org ISWC 2017] in Vienna, Austria on 21st or 22nd of October 2017.

=== Social Dinner ===
Information about a social dinner will be announced at the workshop.

== Program ==

TBA

== Call for Papers - Topics ==
Submission of research papers (including position papers) is via the [https://easychair.org/conferences/?conf=wop2017 WOP 2017 EasyChair] page. Detailed submission instructions can be found on the [[WOP:2017/Submission | WOP 2017 submission page]].

The main aim of the workshop is to discuss and collect solutions to recurrent problems that matter to researchers and practitioners of the Semantic Web field, and that impact on design and engineering of ontologies, linked data, knowledge extraction, and other semantic web applications. We invite the submission of original research results related to the focus areas of the workshop. '''Research papers (maximum 15 pages LNCS style)''' should present mature work and document established results. '''Short papers (maximum 5 pages LNCS style)''' may present proposed research directions, novel ideas, or more general positions or discussions.

'''This year we particularly welcome contributions considering how one can develop high-quality ontologies in general, with or without the help of ODPs.''' Nevertheless, as usual we also welcome
pattern descriptions of all sorts, including patterns geared towards applications in specific domains,
such as geosciences, biomedical science, as well as digital humanities. The following shows main
topics of interest:

* Ontology design patterns, analyses of ODP use, and analyses of pattern-based ontologies
* Correspondence patterns for ontology matching and integration
* Knowledge patterns and knowledge reengineering based on patterns
* Antipatterns and their relations to ODPs
* Methods and tools for developing high-quality ontologies, including ontology engineering by domain experts *new*
* Analyses of quality attributes in ontologies and ontology engineering, and quality assurance approaches for ontology engineering *new*
* Pattern-based ontology design and knowledge engineering
* Pattern-based methods and methodologies for development of semantic applications
* Ontology pattern extraction
* Pattern-based information extraction, ontology learning, and relation to NLP
* Pattern-based ontology evaluation and selection
* Patterns and Linked data (usage, emerging patterns, pattern-driven data publishing, etc.)
* Quality assurance for Linked Data vocabularies *new*
* Patterns for streaming data and evolving knowledge, as well as processes and services
* Data mining patterns and Semantic Web design
* Web semantics from a pattern perspective
* Reasoning with, or using, patterns
* Patterns in semantic social networks, semantic wikis, semantic blogs, as well as in related fields such as business intelligence and conceptual modelling
* ODPs development and use in specific domains including geosciences, life sciences, digital humanities, cultural heritage, e-history, etc.

== Call for Patterns ==
Submissions should be made via the [https://easychair.org/conferences/?conf=wop2017 WOP 2017 Easy Chair] page. Detailed instructions for submitting your pattern can be found at the [[WOP:2017/Submission | WOP 2017 submission page]].

We invite the submission of research results in the form of ontology design patterns (ODPs).
Patterns submitted should have a general relevance to the ontology engineering field, or specific interest within a knowledge domain. Patterns should solve some particular modelling problem, and be of significant interest for discussion at the workshop. Patterns should be original, in the sense that they are the intellectual product of the author(s), however they may still be based on the collective experience of a community.

Pattern submissions for the pattern session will be collected:

* through the [http://ontologydesignpatterns.org/wiki/Submissions:SubmitAPattern ODP portal] and
* by submitting a description of the pattern (pattern description) via EasyChair.

Detailed instructions for patterns submission, including how to submit via the ontologydesignpatterns.org portal, are found at the [[WOP:2017/Submission | submission page]]. Note that an account in the ODP portal is needed for submitting patterns; thus, authors should take care to request an account at least one week before their intended submission.

Pattern submissions can be made in any type of ODPs. Currently, portal templates for submission are provided for the following types of patterns (see general [[OPTypes|typology]] for explanation of the types):

* Content patterns
* Structural patterns: logical and architecture patterns.
* Correspondence patterns: re-engineering and alignment patterns.

For other types of patterns, the author is welcome to submit only a pattern description.

== Submission and Important Dates ==
Submissions should be made via the [https://easychair.org/conferences/?conf=wop2017 WOP 2017 Easy Chair] page. Submissions should be in the Springer Lecture Notes in Computer Science (LNCS) format. [ftp://ftp.springer.de/pub/tex/latex/llncs/latex2e/llncs2e.zip Latex] and [ftp://ftp.springer.de/pub/tex/latex/llncs/word/splnproc1703.zip Word] templates are available. Detailed submission requirements can be found on the [[WOP:2017/Submission | WOP 2017 submission page]].

'''Important dates'''
* Abstract submission (required): July 18th, 2017
* Submission date: July 21st, 2017
* Author notifications: August 24th, 2017
* Camera-ready papers: September 11th, 2017
* Workshops will be held on: 21st or 22nd of October, 2017 (TBD)

=== Best Poster Award ===

The best poster award for WOP 2017 is awarded based on voting by the workshop chairs (excluding those workshop chairs who themselves authored or co-authored a poster). The best poster award takes into account criteria such as the quality of the submission, the relevance and significance of patterns or pattern usages presented, the presentation of the poster during the session, and the level of involvement of the author during the revision phase and poster session discussions during the workshop.

==Proceedings==
Accepted papers and pattern descriptions will be made available on this page.

We will also publish workshop proceedings with CEUR before the workshop.

== WOP Organisation ==

==== WOP2017 Chairs ====

* Chairs:
** [http://www.ida.liu.se/~evabl45 Eva Blomqvist], Linköping University, Sweden (general chair)
** [http://mayor2.dia.fi.upm.es/oeg-upm/index.php/en/teachers/11-ocorcho/ Oscar Corcho], Ontology Engineering Group, Universidad Politecnica de Madrid, Spain (papers co-chair)
** [http://web.stanford.edu/~horridge/ Matthew Horridge], Stanford Center for Bio-Medical Informatics Research, Stanford University, US (papers co-chair)
** [http://www.rinkehoekstra.nl Rinke Hoekstra], Elsevier B.V., the Netherlands (patterns co-chair)
** [https://iccl.inf.tu-dresden.de/web/David_Carral/en David Carral], International Center For Computational Logic, TU Dresden, Germany (patterns co-chair)

For general inquiries, please contact [http://www.ida.liu.se/~evabl45 Eva Blomqvist]

==== Steering Committee ====

The WOP Steering committee - the [http://ontologydesignpatterns.org/wiki/ODPA Board of the Association of Ontology Design and Patterns] - consists of:

Regular members:

* Eva Blomqvist, Linköping University (chair)
* Pascal Hitzler, Wright State University (vice chair)
* Valentina Presutti, ISTC-CNR (vice chair)
* Aldo Gangemi, ISTC-CNR
* Krzysztof Janowicz, University of California, Santa Barbara
* Agnieszka Lawrynowicz, Poznan University of Technology
* Adila Krisnadhi, Wright State University

Members with special appointments:

* Andrea Nuzzolese, ISTC-CNR
* Karl Hammar, Jönköping University

==== Program Committee ====

TBA

[[Category:Event]]
[[Category:WOP]]
{{Event
|Name=WOP 2017
|HasStartDate=2017/10/21
|HasEndDate=2017/10/22
}}
{{Submission deadline
|deadline=2016/07/21
|expired=TBA
|intime=TBA
}}

Reviews:RinkeHoekstra about Object with states

2013-08-12T13:56:11Z

RinkeHoekstra: New page: {{Content OP Proposal Review Template |CreationDate=2013/8/12 |SubmittedBy=RinkeHoekstra |ContentOPUnderReview=Object with states |RevisionID=11667 |Score=0 - needs major revision |ReviewS...

{{Content OP Proposal Review Template
|CreationDate=2013/8/12
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=Object with states
|RevisionID=11667
|Score=0 - needs major revision
|ReviewSummary=The design pattern is intended to allow the representation of objects with states. To this end, the author proposes to define objects as classes with certain properties, and link them to separate individuals that represent the state of the object using a 'hasState' property.

The author does not really discuss related work, most prominently e.g. the perdurants of the DOLCE ontology. Also, I do not think the solution is satisfactory, or at least it is not very generalizable nor very strictly defined. Detailed comments below:

* It seems ontologically strange to define 'objects' as classes, whereas the states are individuals associated to *instances* of those object classes using the 'hasState' property. A more pragmatic solution would be to decide that all individuals of type 'Object' are also states. One can then individuate an object through time (i.e. through all its states) by chaining these states together using a property of choice. Another option would be to consider all states that are instances of a certain object class to be states of the same object. The object class then represents the object as it exists through time (perhaps this is closest to the currently proposed pattern)
* Secondly, it could be argued that there should not be a strict requirement that the states of an object are distinct individuals. For instance, I can be both married and brown-haired at the same time. Distinct individuals is a design decision that the author should argue for explicitly.
* The design pattern does not ensure that individual states are only the state of a *single* object. I.e. the has-state property should be inverse functional.
|ReviewConfidence=fairly high
|ReviewProblems=* It seems ontologically strange to define 'objects' as classes, whereas the states are individuals associated to *instances* of those object classes using the 'hasState' property. A more pragmatic solution would be to decide that all individuals of type 'Object' are also states. One can then individuate an object through time (i.e. through all its states) by chaining these states together using a property of choice. Another option would be to consider all states that are instances of a certain object class to be states of the same object. The object class then represents the object as it exists through time (perhaps this is closest to the currently proposed pattern)
* Secondly, it could be argued that there should not be a strict requirement that the states of an object are distinct individuals. For instance, I can be both married and brown-haired at the same time. Distinct individuals is a design decision that the author should argue for explicitly.
* The design pattern does not ensure that individual states are only the state of a *single* object. I.e. the has-state property should be inverse functional.
|ReviewRelevance=The pattern, and its subject, is very relevant for the community as it is a very often occurring pattern in ontology engineering and modeling in general.
|ReviewBestPractice=The pattern does not really take into account the best practices around this subject. Objects and their states are discussed at length in the literature (though mostly philosophically inspired). It is fair enough that the author takes a pragmatic approach (which is laudable), but consulting related work would most likely improve the quality of the pattern.
|ReviewReusability=The pattern is reusable.
|ReviewUnderstandability=The pattern is understandable (the problem it tries to solve is clear), but it is formally underspecified.
|ReviewClearProblem=The problem is clear
|ReviewClearRelevance=Consequences are not really made clear in the description. For instance, the assumption that states are distinct individuals is not sufficiently motivated nor discussed.
|ReviewFigures=Yes
|ReviewMissing=A discussion of design decisions with reference to the literature would really help engineers in choosing this pattern.
}}

Reviews:RinkeHoekstra about PeriodicInterval

2012-08-21T12:37:18Z

RinkeHoekstra: New page: {{Content OP Proposal Review Template |CreationDate=2012/8/21 |SubmittedBy=RinkeHoekstra |ContentOPUnderReview=PeriodicInterval |RevisionID=11152 |Score=0 - needs major revision |ReviewSum...

{{Content OP Proposal Review Template
|CreationDate=2012/8/21
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=PeriodicInterval
|RevisionID=11152
|Score=0 - needs major revision
|ReviewSummary=The pattern for periodic intervals is highly relevant in many scenarios. It is a useful extension of the OWL Time ontology. It adequately imports the CP annotation schema, but does not really provide any documentation

However, the pattern is too simplistic in its current form. In a way, the current definition can be read as a recipe for instantiating the sub intervals. This is done via a single object property (hasIntervalDurationPerPeriod), but could have been represented in OWL as restrictions on the PeriodicInterval class itself, e.g. it could be defined as a class with only subintervals of a particular length. The authors could spend some time thinking of how to augment their definition to make it more expressive (e.g. how to make sure the subintervals do not overlap)

Furthermore, the pattern does not adequately restrict the class, it only defines domain and range of two new properties. A PeriodicInterval should at least have two subintervals and a period that separates them.
|ReviewConfidence=High
|ReviewProblems=The pattern is overly simplistic (see summary)
|ReviewRelevance=High
|ReviewBestPractice=The pattern does not present a best practice
|ReviewReusability=High
|ReviewRelations=Many in the category Time, but in particular [[Submissions:TimeInterval]] and [[Submissions:Sequence]]
|ReviewUnderstandability=Ok
|ReviewClearProblem=Ok
|ReviewClearRelevance=Ok
|ReviewFigures=Yes
}}

Reviews:RinkeHoekstra about OOPMetrics

2012-08-21T11:51:11Z

RinkeHoekstra: New page: {{Content OP Proposal Review Template |CreationDate=2012/8/21 |SubmittedBy=RinkeHoekstra |ContentOPUnderReview=OOPMetrics |RevisionID=11169 |Score=-1 - reject |ReviewSummary=The ontology h...

{{Content OP Proposal Review Template
|CreationDate=2012/8/21
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=OOPMetrics
|RevisionID=11169
|Score=-1 - reject
|ReviewSummary=The ontology has very limited motivation, and it is unclear to me how the ontology would form a reusable pattern across different ontologies. The idea of ontology design patterns is that they present fragments of ontologies that are common across multiple domains. An ontology for evaluating OOP, however useful, is not a pattern in that sense, it is a very small ontology in its own right, that has a very specific purpose. Perhaps if this ontology could be generalized to rating/ranking/measuring things in general, it could be a pattern. However, this will require some more substantial backing.
|ReviewConfidence=High
|ReviewBestPractice=It is an ontology for measuring best practices in OOP, which is good. However, the ontology itself does not follow best practices (lack of labels, limited documentation etc.
|ReviewReusability=Reusable but only for a very specific task.
|ReviewRelations=None
|ReviewUnderstandability=Ok
|ReviewClearProblem=Far to limited discussion
|ReviewClearRelevance=Relevance is clear for OOP metrics, but not as a pattern.
|ReviewFigures=Yes
}}

Submissions:N-Ary Relation Pattern (OWL 2)

2010-10-13T17:02:58Z

RinkeHoekstra:

{{Logical_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Nary-relation.png
}}
{{Logical OP General Template
|Name=N-Ary Relation Pattern (OWL 2)
|SubmittedBy=RinkeHoekstra
|Author=Rinke Hoekstra
}}
{{Logical OP Description Template
|Motivation=OWL does not support N-Ary relations. This means that an n-ary relation can only be represented in its reified form in OWL. This is problematic, as the relational character is then completely lost.
|Aim=The aim of this pattern is to allow the inference of property relations between the different relata of the original N-Ary relation based on its reification.
|Solution=The N-Ary relation is reified by creating a class for the relation (NR), and creating properties and classes for the domain (D) and ranges (R1-Rn) of the relation (that is, if the relation is directional). The NR class is specified using a local reflexivity restriction of the form: NR equiv is_NR some Self. We then specify role chains for each of the binary relations between the domain and ranges. For instance: has_NR o is_NR o r1 -> has_r1
|Elements=OWL 2 local reflexivity and property chains.
|Implementation=Alter the axioms in the reusable component to suit your needs. The component contains an N-Ary relation with four ranges and a domain. This can easily be extended or reduced.
|ReusableComponent=http://purl.org/net/nary-relation
|ComponentType=OWL 2 RDF/XML
}}
{{Logical OP Example Template
|ProblemExample=See the picture, where the cancer diagnosis relation of the SWBP N-Ary relations patterns is expressed using this design pattern.
|Consequences=The patient (Christine, P) is related both to the diagnosis value (Cancer) and its probability.
}}
{{Logical OP Reference Template}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)

In your academic publications, please refer to this pattern by citing:
Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.

[[Category:Waiting for review]]
{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}

Submissions:N-Ary Relation Pattern (OWL 2)

2010-10-13T12:58:26Z

RinkeHoekstra: Article is waiting for review.

{{Logical_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Nary-relation.png
}}
{{Logical OP General Template
|Name=N-Ary Relation Pattern (OWL 2)
|SubmittedBy=Rinke Hoekstra
|Author=Rinke Hoekstra
}}
{{Logical OP Description Template
|Motivation=OWL does not support N-Ary relations. This means that an n-ary relation can only be represented in its reified form in OWL. This is problematic, as the relational character is then completely lost.
|Aim=The aim of this pattern is to allow the inference of property relations between the different relata of the original N-Ary relation based on its reification.
|Solution=The N-Ary relation is reified by creating a class for the relation (NR), and creating properties and classes for the domain (D) and ranges (R1-Rn) of the relation (that is, if the relation is directional). The NR class is specified using a local reflexivity restriction of the form: NR equiv is_NR some Self. We then specify role chains for each of the binary relations between the domain and ranges. For instance: has_NR o is_NR o r1 -> has_r1
|Elements=OWL 2 local reflexivity and property chains.
|Implementation=Alter the axioms in the reusable component to suit your needs. The component contains an N-Ary relation with four ranges and a domain. This can easily be extended or reduced.
|ReusableComponent=http://purl.org/net/nary-relation
|ComponentType=OWL 2 RDF/XML
}}
{{Logical OP Example Template
|ProblemExample=See the picture, where the cancer diagnosis relation of the SWBP N-Ary relations patterns is expressed using this design pattern.
|Consequences=The patient (Christine, P) is related both to the diagnosis value (Cancer) and its probability.
}}
{{Logical OP Reference Template}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)

In your academic publications, please refer to this pattern by citing:
Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.
{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}
[[Category:Waiting for review]]

Submissions:Roles (OWL 2)

2010-10-13T12:58:02Z

RinkeHoekstra: Article is waiting for review.

{{Content_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Role.png
}}
{{Content OP Proposal Template
|SubmittedBy=Rinke Hoekstra
|Name=Roles (OWL 2)
|ContentODPAlsoKnownAs=Roles, Functions, Actions and Agents (OWL 2)
|Intent=Capture the relational nature of roles using OWL 2 property chains, using the[[Community:Social Reality (OWL 2)|Social Reality (OWL 2)]] pattern. Using this pattern the role-as-relation and role-as-class approach can coexist peacefully inside a single ontology.
|ContentODPDescription=Extend the[[Community:Social Reality (OWL 2)|Social Reality (OWL 2)]] pattern with classes for roles, functions, agents and actions. Create a new subproperty "plays" for the counts-as relation. We create new property chains to traverse the context and plays relation, that allow us to infer the role-as-relation from the role-as-class description.

Imports the[[Community:Social Reality (OWL 2)|Social Reality (OWL 2)]] pattern from [http://purl.org/net/social-reality].
|ReusableOWLBuildingBlock=http://purl.org/net/roles
}}
{{Element list header}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)

In your academic publications, please refer to this pattern by citing:

Rinke Hoekstra. Representing social reality in OWL 2. In Evren Sirin and Kendall Clark, editors, Proceedings of OWLED 2010, June 2010.

and/or

Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.

{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}
[[Category:Waiting for review]]

File:Role.png

2010-10-13T12:57:23Z

RinkeHoekstra: uploaded a new version of "Image:Role.png"

Roles, functions, thematic roles (agent action role)

File:Role.png

2010-10-13T12:45:32Z

RinkeHoekstra: uploaded a new version of "Image:Role.png"

Roles, functions, thematic roles (agent action role)

Submissions:Roles (OWL 2)

2010-10-13T12:36:45Z

RinkeHoekstra:

Submissions:Roles (OWL 2)

2010-10-13T12:35:55Z

RinkeHoekstra: New page: {{Content_OP_Proposal_toolbar}} {{Graphical representation header}} {{Graphical representation |ImageName=Role.png }} {{Content OP Proposal Template |SubmittedBy=Rinke Hoekstra |Name=Roles...

File:Role.png

2010-10-13T12:30:40Z

RinkeHoekstra: Roles, functions, thematic roles (agent action role)

Roles, functions, thematic roles (agent action role)

Submissions:Social Reality (OWL 2)

2010-10-13T12:20:16Z

RinkeHoekstra:

{{Content_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Relative-facts.png
}}
{{Content OP Proposal Template
|SubmittedBy=Rinke Hoekstra
|Name=Social Reality (OWL 2)
|ContentODPAlsoKnownAs=Observer Relative and Institutional Facts
|Intent=Capture Searle's theory on observer relative and institutional facts, used in creating social reality. See also the OWL 2 Agent-Role pattern.
|ContentODPDescription=Using the approach taken in [[Submissions:N-Ary_Relation_Pattern_(OWL_2)]] we place an observer relative fact (OR) in a context (the observer, C) and the brute fact (BF). For the theory This is a reified relation, and we then specify a role chain that allows us to directly relate the observer (context C) to the brute fact using a relation that reflects the fact that the brute fact counts-as some observer relative fact. See Searle, 1995 (construction of social reality) and/or:

Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.

Rinke Hoekstra. Representing social reality in owl 2. In Evren Sirin and Kendall Clark, editors, Proceedings of OWLED 2010, June 2010.

for the details of this theory.
|ReusableOWLBuildingBlock=http://purl.org/net/social-reality
|KnownUse=http://purl.org/net/lkif-core
}}
{{Element list header}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)

In your academic publications, please refer to this pattern by citing:

Rinke Hoekstra. Representing social reality in OWL 2. In Evren Sirin and Kendall Clark, editors, Proceedings of OWLED 2010, June 2010.

and/or

Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.
{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}

Submissions:Social Reality (OWL 2)

2010-10-13T11:00:31Z

RinkeHoekstra: New page: {{Content_OP_Proposal_toolbar}} {{Graphical representation header}} {{Graphical representation |ImageName=Relative-facts.png }} {{Content OP Proposal Template |SubmittedBy=Rinke Hoekstra |...

{{Content_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Relative-facts.png
}}
{{Content OP Proposal Template
|SubmittedBy=Rinke Hoekstra
|Name=Social Reality (OWL 2)
|ContentODPAlsoKnownAs=Observer Relative and Institutional Facts
|Intent=Capture Searle's theory on observer relative and institutional facts, used in creating social reality. See also the OWL 2 Agent-Role pattern.
|ContentODPDescription=Using the approach taken in [Submissions:N-Ary_Relation_Pattern_(OWL_2)] we place an observer relative fact (OR) in a context (the observer, C) and the brute fact (BF). For the theory This is a reified relation, and we then specify a role chain that allows us to directly relate the observer (context C) to the brute fact using a relation that reflects the fact that the brute fact counts-as some observer relative fact. See Searle, 1995 (construction of social reality) and/or:

Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.

Rinke Hoekstra. Representing social reality in owl 2. In Evren Sirin and Kendall Clark, editors, Proceedings of OWLED 2010, June 2010.

for the details of this theory.
|ReusableOWLBuildingBlock=http://purl.org/net/social-reality
|KnownUse=In the LKIF-Core ontology
|CPInstantiationExample=See the building block file
}}
{{Element list header}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)

In your academic publications, please refer to this pattern by citing:

Rinke Hoekstra. Representing social reality in OWL 2. In Evren Sirin and Kendall Clark, editors, Proceedings of OWLED 2010, June 2010.

and/or

Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.
{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}

File:Relative-facts.png

2010-10-13T10:51:23Z

RinkeHoekstra: Design pattern for relative facts (observer relative, institutional) based on Searle, 1995

Design pattern for relative facts (observer relative, institutional) based on Searle, 1995

Submissions:N-Ary Relation Pattern (OWL 2)

2010-10-13T10:47:57Z

RinkeHoekstra:

{{Logical_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Nary-relation.png
}}
{{Logical OP General Template
|Name=N-Ary Relation Pattern (OWL 2)
|SubmittedBy=Rinke Hoekstra
|Author=Rinke Hoekstra
}}
{{Logical OP Description Template
|Motivation=OWL does not support N-Ary relations. This means that an n-ary relation can only be represented in its reified form in OWL. This is problematic, as the relational character is then completely lost.
|Aim=The aim of this pattern is to allow the inference of property relations between the different relata of the original N-Ary relation based on its reification.
|Solution=The N-Ary relation is reified by creating a class for the relation (NR), and creating properties and classes for the domain (D) and ranges (R1-Rn) of the relation (that is, if the relation is directional). The NR class is specified using a local reflexivity restriction of the form: NR equiv is_NR some Self. We then specify role chains for each of the binary relations between the domain and ranges. For instance: has_NR o is_NR o r1 -> has_r1
|Elements=OWL 2 local reflexivity and property chains.
|Implementation=Alter the axioms in the reusable component to suit your needs. The component contains an N-Ary relation with four ranges and a domain. This can easily be extended or reduced.
|ReusableComponent=http://purl.org/net/nary-relation
|ComponentType=OWL 2 RDF/XML
}}
{{Logical OP Example Template
|ProblemExample=See the picture, where the cancer diagnosis relation of the SWBP N-Ary relations patterns is expressed using this design pattern.
|Consequences=The patient (Christine, P) is related both to the diagnosis value (Cancer) and its probability.
}}
{{Logical OP Reference Template}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)

In your academic publications, please refer to this pattern by citing:
Rinke Hoekstra. Ontology Representation – Design Patterns and Ontologies that Make Sense, volume 197 of Frontiers of Artificial Intelligence and Applications. IOS Press, Amsterdam, June 2009.
{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}

Submissions:N-Ary Relation Pattern (OWL 2)

2010-10-13T10:35:53Z

RinkeHoekstra:

{{Logical_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Nary-relation.png
}}
{{Logical OP General Template
|Name=N-Ary Relation Pattern (OWL 2)
|SubmittedBy=Rinke Hoekstra
|Author=Rinke Hoekstra
}}
{{Logical OP Description Template
|Motivation=OWL does not support N-Ary relations. This means that an n-ary relation can only be represented in its reified form in OWL. This is problematic, as the relational character is then completely lost.
|Aim=The aim of this pattern is to allow the inference of property relations between the different relata of the original N-Ary relation based on its reification.
|Solution=The N-Ary relation is reified by creating a class for the relation (NR), and creating properties and classes for the domain (D) and ranges (R1-Rn) of the relation (that is, if the relation is directional). The NR class is specified using a local reflexivity restriction of the form: NR equiv is_NR some Self. We then specify role chains for each of the binary relations between the domain and ranges. For instance: has_NR o is_NR o r1 -> has_r1
|Elements=OWL 2 local reflexivity and property chains.
|Implementation=Alter the axioms in the reusable component to suit your needs. The component contains an N-Ary relation with four ranges and a domain. This can easily be extended or reduced.
|ReusableComponent=http://purl.org/net/nary-relation
|ComponentType=OWL 2 RDF/XML
}}
{{Logical OP Example Template
|ProblemExample=See the picture, where the cancer diagnosis relation of the SWBP N-Ary relations patterns is expressed using this design pattern.
|Consequences=The patient (Christine, P) is related both to the diagnosis value (Cancer) and its probability.
}}
{{Logical OP Reference Template}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)
{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}

Submissions:N-Ary Relation Pattern (OWL 2)

2010-10-13T10:33:59Z

RinkeHoekstra: New page: {{Logical_OP_Proposal_toolbar}} {{Graphical representation header}} {{Graphical representation |ImageName=Image:Nary-relation.png }} {{Logical OP General Template |Name=N-Ary Relation Patt...

{{Logical_OP_Proposal_toolbar}}
{{Graphical representation header}}
{{Graphical representation
|ImageName=Image:Nary-relation.png
}}
{{Logical OP General Template
|Name=N-Ary Relation Pattern (OWL 2)
|SubmittedBy=Rinke Hoekstra
|Author=Rinke Hoekstra
}}
{{Logical OP Description Template
|Motivation=OWL does not support N-Ary relations. This means that an n-ary relation can only be represented in its reified form in OWL. This is problematic, as the relational character is then completely lost.
|Aim=The aim of this pattern is to allow the inference of property relations between the different relata of the original N-Ary relation based on its reification.
|Solution=The N-Ary relation is reified by creating a class for the relation (NR), and creating properties and classes for the domain (D) and ranges (R1-Rn) of the relation (that is, if the relation is directional). The NR class is specified using a local reflexivity restriction of the form: NR equiv is_NR some Self. We then specify role chains for each of the binary relations between the domain and ranges. For instance: has_NR o is_NR o r1 -> has_r1
|Elements=OWL 2 local reflexivity and property chains.
|Implementation=Alter the axioms in the reusable component to suit your needs. The component contains an N-Ary relation with four ranges and a domain. This can easily be extended or reduced.

|ReusableComponent=http://purl.org/net/nary-relation
|ComponentType=OWL 2 RDF/XML
}}
{{Logical OP Example Template
|ProblemExample=See the picture, where the cancer diagnosis relation of the SWBP N-Ary relations patterns is expressed using this design pattern.
|Consequences=The patient (Christine, P) is related both to the diagnosis value (Cancer) and its probability.
}}
{{Logical OP Reference Template}}
{{Additional information header}}
This pattern is made available under the CC-BY licence (http://creativecommons.org/licenses/by/2.0/). Under this licence the pattern can be freely reused but reusers should include an attribution to the original author (Rinke Hoekstra)
{{Scenarios about me}}
{{Reviews about me}}
{{Modeling issues about me}}
{{My references}}

File:Nary-relation.png

2010-10-13T09:57:56Z

RinkeHoekstra: Diagram and example for an N-Ary Relation pattern in OWL 2

Diagram and example for an N-Ary Relation pattern in OWL 2

File:Nary-relation.pdf

2010-10-13T09:55:45Z

RinkeHoekstra: Diagram and example of an N-ary relation pattern

Diagram and example of an N-ary relation pattern

Community:RinkeHoekstra about Define Hybrid Class Resolving Disjointness due to Subsumption 2

2009-10-25T16:10:33Z

RinkeHoekstra: New page: {{Content OP Proposal Review Template |CreationDate=2009-10-25 |SubmittedBy=RinkeHoekstra |ContentOPUnderReview=Define Hybrid Class Resolving Disjointness due to Subsumption |RevisionID=58...

{{Content OP Proposal Review Template
|CreationDate=2009-10-25
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=Define Hybrid Class Resolving Disjointness due to Subsumption
|RevisionID=5867
|Score=1 - needs minor revision
|ReviewSummary=Correction to my previous review (considering the discussion at WOP at ISWC 2009). Unfortunately I cannot change it anymore, some of my remarks were slightly jet-lagged induced.

}}

2009-10-25T10:45:53Z

RinkeHoekstra:

<snip> (was silly and jet-lag induced)

[[User:RinkeHoekstra|RinkeHoekstra]] 02:17, 25 October 2009 (UTC)

2009-10-25T02:17:55Z

RinkeHoekstra: Added review

I am unable to post an open review (There's something amiss with the list) so I will post my review here:

The authors have not considered the awkwardness of RDF semantics. The problem is that RDF reificaiton only works in one direction, i.e. every triple can be reified as an rdf:Statement with rdf:subject rdf:predicate rdf:object (that is, for every triple such a statement exists). However, it does not work in the other direction: the existence of an rdf:Statement does not imply the existence of the triple.

This may certainly cause unexpected behaviour with the proposed pattern. Let alone the fact that the proposed solution brings the ontology in OWL Full which is not ideal for many situations.

[[User:RinkeHoekstra|RinkeHoekstra]] 02:17, 25 October 2009 (UTC)

Community:RinkeHoekstra about Partition

2009-10-25T02:07:39Z

RinkeHoekstra: New page: {{Content OP Proposal Review Template |CreationDate=2009-10-25 |SubmittedBy=RinkeHoekstra |ContentOPUnderReview=Partition |RevisionID=5901 |Score=0 - needs major revision |ReviewSummary=We...

{{Content OP Proposal Review Template
|CreationDate=2009-10-25
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=Partition
|RevisionID=5901
|Score=0 - needs major revision
|ReviewSummary=Well.. what can I say. This pattern again (cf. the pattern for negative property assertions) describes what is already covered by the OWL 2 specification as syntactic sugar. Although I agree that this may be useful for those not willing or able to commit to OWL 2, the pattern should be attributed to the OWL 2 specification, and not the current author of this pattern.

}}

Community:RinkeHoekstra about Define Hybrid Class Resolving Disjointness due to Subsumption

2009-10-25T02:02:00Z

{{Content OP Proposal Review Template
|CreationDate=2009-10-25
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=Define Hybrid Class Resolving Disjointness due to Subsumption
|RevisionID=5867
|Score=-1 - reject
|ReviewSummary=The reason I propose to reject this pattern is that it is decidedly silly.

The authors say that "The definition of the Hybrid Class is the union (OR) of the definitions of the disjoint classes."

The reason is that the two classes A and B are subsumed by their union A V B. What the pattern does is that we leave implicit what the '''proper''' type of the individual is: we simply do not know whether the individual is in A or B.
However, any individual instance of A V B will always be either in A or in B, but '''never''' in both (since this is what disjointness specifies).

The example therefore does not applies. If Animal_Plant is defined as the union of Animal and Plant (both disjoint) then there still exist no Animal_Plant instances... and this is what the pattern supposedly tries to solve.

}}

Community:RinkeHoekstra about OnlynessIsLoneliness (OIL)

2009-10-24T20:15:47Z

RinkeHoekstra: New page: {{Content OP Proposal Review Template |CreationDate=2009-10-24 |SubmittedBy=RinkeHoekstra |ContentOPUnderReview=OnlynessIsLoneliness (OIL) |RevisionID=5787 |Score=0 - needs major revision ...

{{Content OP Proposal Review Template
|CreationDate=2009-10-24
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=OnlynessIsLoneliness (OIL)
|RevisionID=5787
|Score=0 - needs major revision
|ReviewSummary=I am not sure in what sense this is a design pattern, or rather a strategy to debug ontologies.

More importantly, the authors fail to note the correspondence with the issue with domain and range definitions of properties where multiple domain and range assertions typically result in equivalence between the classes occurring in the assertions. Lastly, this equivalence is really the problem with the allvaluesfrom restriction as well.

Nice title though.

}}

Community:RinkeHoekstra about NegativePropertyAssertions

2009-10-24T20:04:46Z

RinkeHoekstra: New page: {{Content OP Proposal Review Template |CreationDate=2009-10-24 |SubmittedBy=RinkeHoekstra |ContentOPUnderReview=NegativePropertyAssertions |RevisionID=5903 |Score=-1 - reject |ReviewSummar...

{{Content OP Proposal Review Template
|CreationDate=2009-10-24
|SubmittedBy=RinkeHoekstra
|ContentOPUnderReview=NegativePropertyAssertions
|RevisionID=5903
|Score=-1 - reject
|ReviewSummary=The author of this pattern fails to mention that the negative property assertions of OWL 2 are purely syntactic sugar for exactly the pattern he describes.

They are explicitly introduced in OWL 2 in e.g. the new features and rationale document.

In conclusion, this pattern may be useful to users sticking with OWL 1 (although I have no idea why they would want to), and as such can be a valuable contribution to this wiki. However, it would be wrong to attribute it to the current author (though he should receive credit for making it available on the wiki).

[1] http://www.w3.org/TR/2009/PR-owl2-new-features-20090922/#F3:_NegativeObjectPropertyAssertion_and_NegativeDataPropertyAssertion
}}