FOIS 2014 - Notes on the Technical Sessions - Day 1
Michael Gruninger
A Sideways Look at Upper OntologiesMichael Gruninger, Torsten Hahmann, Megan Katsumi and Carmen Chui
Method based on axiomatization and proof to compare and connect diferente Upper Ontologies.
1) How can we understand to what ontological commitments an Upper Ontology commits ?
This is done by decomposing the Upper ontology in a set of generic ontologies which are its modules. E.g. ontology on Betweeness, Timepoint ontology
He claimed they proved (not shown here) that DOLCE is reduceable to a set of more traditional mathematical ontologies.
There is a base in his lab that puts together these different ontologies. He showed some excerpts represented by graphs. For instance, different perspectives on Betweeness, Timepoint Ontologies.
Ontological commitments vs. Ontological choices
*ask his slides (very interesting distinction!)
Every generic ontology has a set of commitments and a set of choices (the choices are materialized by axioms that constrain the ontology).
2) How can an upper ontology be partially reused?
He presents a model (kind of architectural model), which he calls sideways view, providing a kind of method of how this can be done. But for me, it was too fast to understand.
3) How can we partially integrate an upper ontology that agrees with only...
Identifying relationships between two distinct ontological repositories.
When there are no mappings between two upper ontologies, he uses a reference model based on mathematical ontologies.
4) How can we integrate multiple extensions of an upper ontology when they are mutually inconsistente which each other?
Generalized Differences - capture the ontological commitments and choices which are made by one but not by the other ontology. It is a set of entailments that are missing in one of the ontologies.
This is important to understand to which limits we can share.
Must the ontologies be heavily formalized? Just enough, he answered. The more axioms the better but you can already reach some mapping with incomplete formalization, because usually, there needs to be some room for flexibility (i.e. the ontology should allow different models).
How much does the formalism commitment influence their view?
Michael Gruninger
A Sideways Look at Upper OntologiesMichael Gruninger, Torsten Hahmann, Megan Katsumi and Carmen Chui
Method based on axiomatization and proof to compare and connect diferente Upper Ontologies.
1) How can we understand to what ontological commitments an Upper Ontology commits ?
This is done by decomposing the Upper ontology in a set of generic ontologies which are its modules. E.g. ontology on Betweeness, Timepoint ontology
He claimed they proved (not shown here) that DOLCE is reduceable to a set of more traditional mathematical ontologies.
There is a base in his lab that puts together these different ontologies. He showed some excerpts represented by graphs. For instance, different perspectives on Betweeness, Timepoint Ontologies.
Ontological commitments vs. Ontological choices
*ask his slides (very interesting distinction!)
Every generic ontology has a set of commitments and a set of choices (the choices are materialized by axioms that constrain the ontology).
2) How can an upper ontology be partially reused?
He presents a model (kind of architectural model), which he calls sideways view, providing a kind of method of how this can be done. But for me, it was too fast to understand.
3) How can we partially integrate an upper ontology that agrees with only...
Identifying relationships between two distinct ontological repositories.
When there are no mappings between two upper ontologies, he uses a reference model based on mathematical ontologies.
4) How can we integrate multiple extensions of an upper ontology when they are mutually inconsistente which each other?
Generalized Differences - capture the ontological commitments and choices which are made by one but not by the other ontology. It is a set of entailments that are missing in one of the ontologies.
This is important to understand to which limits we can share.
Must the ontologies be heavily formalized? Just enough, he answered. The more axioms the better but you can already reach some mapping with incomplete formalization, because usually, there needs to be some room for flexibility (i.e. the ontology should allow different models).
How much does the formalism commitment influence their view?
Nenhum comentário:
Postar um comentário