Elements of the Financial Industry Business ontology (click through for details):

Operational Ontologies

An 'operational ontology' is an RDF/OWL based ontology which is fine tuned for a specific business purpose. This is distinct from the material in the FIBO OMG specifications or the overall conceptual ontology models for the following reasons:

• Business Conceptual Ontology defines the meanings of things in the domain of discourse. In financial securities, meanings are generally grounded in some legal or financial concepts; operational ontologies need only define classes and properties for things that will correspond to instance data;
• OMG specifications will cover all the concepts which may be of relevence in some use case - that is, they cover more than one use case;
• The role of a conceptual model and the role of an application data store are very different - even when, as with semantic models, the architecture and model components may be the same. That is, a conceptual model (including an ontology of business terms), is a technology-neutral view of the business requirements, whereas an operational OWL ontology, having instance data (called OWL Individuals), is a physical datastore. Even if they were the same, their roles are different and so are the constraints and requirements under which they are built.
• A conceptual model may use multiple inheritance (polyhierarchical taxonomies) to classify the subject matter in multiple ways. This is quite correct for a conceptual model but adds processing overhead for no real value in an operational application, be it an ontology or a conventional database application.
• The FIBO conceptual models make extensive use of partitioning of the upper parts of the model. This is essential in disambiguating similar terms (consider for example the difference between a monetary amount as a unit of measure, and an amount of actual money). This partitioning (which is also polyhierarchical) may add some processing overhead to the application, again for no benefit.
• Summary: There are many do's and don'ts when building an operational OWL ontology for a semantics reasoning or other Semantic Web application. This is true of any type of application and there is no reason for it to be less so when building semantic technology applications. However, these technical design constraints should not be carried back to the conceptual ontology itself - otherwise it loses its value as a description of the subject matter.

The EDM Council and OMG Finance Domain Task Force members are working on developing a series of operational ontologies, starting with derivatives and business entity concepts. In the course of doing this, we expect to more formally identify what are the differences between these and conceptual ontologies, and what are the rules and heuristics for extracting or deriving operational ontology material from the conceptual ontologies. At the same time, the conceptual ontologies are developed in a strongly modular fashion, so that for many applications it may be possible (if performance constraints allow it) to simply use a sub-set of the available modules.

In short, this is very much an ongoing research activity at the present time.

Operational ontologies Overview

• RDF/OWL Extraction of FIBO Content
• Focused on specific use case
• Will use only a fraction of the terms in FIBO-BE etc.
• Delivers the benefits of OWL based reasoning, classification etc.
• Remains an accurate (conformant) representation of the business subject matter in RDF/OWL notation.

Operational Ontologies Rationale

Demonstrate potential of Semantic Technology (RDF/OWL) to deliver real business results in applications.

Demonstrates:

• Reasoning capabilities
• Automatic classification – derivatives etc.
• Semantic Querying