Dear Fabian, (01)
> I had another look at Ontohub (the new version) and have two and a half
> questions: (02)
indeed, these questions are more about OntoIOp; that's why I put the
OntoIOp forum into cc. Ontohub just implements what's there in OntoIOp. (03)
> What is the definition of logic vs. language you are using? The way I
> learned it a long time ago, a logic is a formal language + an inference
> relationship (either specified by a semantics or a deductive system).
> However, this seems to be not the way you are using the terms, since
> Common Logic is classified as language. The situation is further
> complicated by the fact that Common Logic and the OWL languages can be
> serialized in different ways. So we seem to have actually 4 different
> kinds of entities: abstract language, abstract logic, concrete language,
> concrete logic. -- Even if you disagree with this distinction, i think
> it would help if you put some short explanation of the terms "Logic" and
> "Language" on the website. (04)
Your terms are named as follows in OntoIOp:
abstract language: ontology language
abstract logic: logic, institute, institution (note that institute and
institution are technical terms with a mathematical definition, and
"logic" is used as a common term for both)
concrete language: serialization
concrete logic: not needed (05)
Here are the definitions from the OntoIOp working draft: (06)
ontology language: language that is used for the formal specification of
ontologies, equipped with a formal, declarative, logic-based semantics,
plus nonlogical annotations (07)
logic: specification of valid reasoning that comprises signatures,
sentences, models, and a satisfaction relation between models and sentences
Note: Most ontology languages have an underlying logic.
Example: SROIQ}(D) is the logic underlying OWL 2 DL. (08)
institute: logical metaframework based on set theory and order theory (09)
institution: logical metaframework based on category theory, also
covering the semantics of renamings and combinations (010)
serialization: specific syntactic encoding of a given ontology language.
Note: Serializations serve as standard formats for exchanging ontologies
between tools.
Examples:
- OWL uses the term “serialization”; the following are standard OWL
serializations: OWL functional-style syntax, OWL XML, OWL Manchester
syntax, plus any serialization of RDF (e.g. RDF/XML, Turtle, …)
- Common Logic uses the term “dialect”; the following are standard
Common Logic dialects: Common Logic Interchange Format (CLIF),
Conceptual Graph Interchange Format (GCIF), eXtended Common Logic Markup
Language (XCL). (011)
> A related question (the half one) is whether mappings are linked to
> logics or languages. It seems that currently in OntoHub mappings are
> linked to both. That seems a little bit odd to me. Let's assume we
> consider three concrete logics: FOL in Prover9 syntax, CLIF, and CLIF*,
> which basically consists of CLIF but with a closed world semantics for
> negation. The mapping from Prover9-FOL to CLIF will differ from the
> mapping from Prover9-FOL to CLIF*, since in the latter case the
> negations need to be treated differently. Shouldn't mappings be always
> be from logics to logics? (Or, more accurately, from abstract logics to
> abstract logics, and from concrete logics to concrete logics?) (012)
Again, I quote the OntoIOp working draft: (013)
ontology language translation: mapping from constructs in the source
ontology language to their equivalents in the target ontology language
Note: An ontology language translation shall satisfy the property that
the result of a translation is a well-formed text in the target language. (014)
logic translation: mapping of a source logic into a target logic
(mapping signatures, sentences and models) that keeps or encodes the
logical content of ontologies (015)
So far the WD.
Note that an ontology language translation typically could also be
specified on the abstract syntaxes, not only on serializations. Ontology
language translations are those that are implemented on a computer.
Logic translations are those being written down in research papers, and
are formalised as institute comorphisms or institution comorphisms
(maybe this should be added to the WD). (016)
So FOL, CLIF, and CLIF* will be three different ontology languages and
also three different logics. Prover9 syntax is a serialisation of FOL. (017)
> My last question is whether there will be a space to upload the mappings
> in some standardized form? It seems that right now the mappings are
> just registered. If you allow users to upload the mappings in some
> standardized fashion, it will be possible to write tools that use these
> mappings to translate from one ontology language into another. (018)
Good point. This is subject of future research - actually the project
OntoLATIN (that I have sent around earlier) will enable exactly this. (019)
All the best,
Till (020)
--
Prof. Dr. Till Mossakowski Cartesium, room 2.51 Phone +49-421-218-64226
DFKI GmbH Bremen Fax +49-421-218-9864226
Cyber-Physical Systems Till.Mossakowski@xxxxxxx
Enrique-Schmidt-Str. 5, D-28359 Bremen
http://www.informatik.uni-bremen.de/~till/ (021)
Deutsches Forschungszentrum fuer Kuenstliche Intelligenz GmbH
principal office, *not* the address for mail etc.!!!:
Trippstadter Str. 122, D-67663 Kaiserslautern
management board: Prof. Wolfgang Wahlster (chair), Dr. Walter Olthoff
supervisory board: Prof. Hans A. Aukes (chair)
Amtsgericht Kaiserslautern, HRB 2313 (022)
_________________________________________________________________
To Post: mailto:ontoiop-forum@xxxxxxxxxxxxxxxx
Message Archives: http://ontolog.cim3.net/forum/ontoiop-forum/
Config/Unsubscribe: http://ontolog.cim3.net/mailman/listinfo/ontoiop-forum/
Community Files (open): http://interop.cim3.net/file/pub/OntoIOp/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntoIOp (023)
|