Accession Number:



A Probabilistic Ontology Development Methodology

Descriptive Note:

Conference paper

Corporate Author:


Report Date:


Pagination or Media Count:



The use of ontologies is on the rise, as they facilitate interoperability and provide support for automation. Today, ontologies are popular in areas such as the Semantic Web, knowledge engineering, Artificial Intelligence and knowledge management. However, many real world problems in these disciplines are burdened by incomplete information and other sources of uncertainty which traditional ontologies cannot represent. Therefore, a means to incorporate uncertainty is a necessity. Probabilistic ontologies extend current ontology formalisms to provide support for representing and reasoning with uncertainty. Traditional ontologies provide a hierarchical structure of entity classes and a formal way of expressing their relationships with first-order expressivity, which supports logical reasoning. However, they lack built-in, principled support to adequately account for uncertainty. Applying simple probability annotations to ontologies fails to convey the structure of the probabilistic representation. Similarly, other less expressive probability schemes do not convey the ontology structure, and are also inadequate. Representation of uncertainty in real-world problems requires probabilistic ontologies, which integrate the inferential reasoning power of probabilistic representations with the first-order expressivity of ontologies. Developing a probabilistic ontology is more complex than simply assigning probability to a class instantiation or representing a probability scheme using ontology constructs. Standard ontological engineering methods provide insufficient support for the complexity of probabilistic ontology development. Therefore, a specific methodology is needed to develop probabilistic ontologies from conceptualization to implementation. We introduce a systematic approach to probabilistic ontology development which focuses on evolving a traditional ontology from conceptualization to probabilistic ontology implementation for real-world problems.

Subject Categories:

  • Information Science

Distribution Statement: