Bayes OWL
From semanticweb.org
BayesOWL or OntoBayes is a probabilistic extension to the Ontology Language OWL. It defines probabilistic relatenes of distinct classes.
As for example we want to say that the probability that a Human is Female is 55.5%. In classic OWL we can't express such a thing. In Bayes OWL this can get expressed.
<prob:CondProbObjeF rdf:ID="P(Female|Human)">
<prob:hasCondition>
<rdf:value>&ont;Human</rdf:value>
</prob:hasCondition>
<prob:hasVariable>
<rdf:value>&ont;Female</rdf:value>
</prob:hasVariable>
<prob:hasProbValue>0.505</prob:hasProbValue>
</prob:CondProbObjeF>
Now, for full clearance, we define also the other probabilities:
<!-- probability for being intersexual -->
<prob:CondProbObjeF rdf:ID="P(Intersexual|Human)">
<prob:hasCondition>
<rdf:value>&ont;Human</rdf:value>
</prob:hasCondition>
<prob:hasVariable>
<owl:Class rdf:ID="Intersexual">
<rdfs:subClassOf>
<owl:unionOf rdf:parseType="Collection">
<owl:Thing rdf:about="&ont;Male"/>
<owl:Thing rdf:about="&ont;Female"/>
</owl:unionOf>
</rdfs:subClassOf>
</owl:Class>
</prob:hasVariable>
<prob:hasProbValue>0.0005</prob:hasProbValue>
</prob:CondProbObjeF>
<!-- probability for being man -->
<prob:CondProbObjeF rdf:ID="P(Male|Human)">
<prob:hasCondition>
<rdf:value>&ont;Human</rdf:value>
</prob:hasCondition>
<prob:hasVariable>
<rdf:value>&ont;Male</rdf:value>
</prob:hasVariable>
<prob:hasProbValue>0.049</prob:hasProbValue>
</prob:CondProbObjeF>
In Table form this looks now like:
| Human | Male | ||
|---|---|---|---|
| true | false | ||
| Female | true | 0.0005 | 0.505 |
| false | 0.049 | NULL | |
Note that in the given Example the sum of all Values is 1 to express 100%. This is not needed to be the case, because one can scale the values. The NULL Value shows that this is not modeled and therefore an impossible state.
[edit] References
- Zhongli Ding, Yun Peng: A Probabilistic Extension to Ontology Language OWL
- Proceedings of the 37th Hawaii International Conference on System Sciences
- University of Maryland Baltimore County 2004
- Zhongli Ding, Yun Peng, Rong Pan: A Bayesian Approach to Uncertainty Modelling in OWL Ontology
- Proceedings of the International Conference on Advances in Intelligent Systems - Theory and Applications
- University of Maryland Baltimore County 2004
- Zhongli Ding, Yun Peng: Modifying Bayesian Networks by Probability Constraints
- Proceedings of the 21st Conference on Uncertainty in Artificial Intelligence
- University of Maryland Baltimore County 2005
- Zhongli Ding, Yun Peng, Rong Pan, Yang Yu: A Bayesian Methodology towards Automatic Ontology Mapping
- Proceedings of the AAAI-05 C&O Workshop on Contexts and Ontologies: Theory, Practice and Applications
- University of Maryland Baltimore County 2005
- Yi Yang, Jacques Calmet, OntoBayes: An Ontology-Driven Uncertainty Model
- University of Karlsruhe (TH) 2005
