gruninger / tlo Goto Github PK

The term ‘ontology’ was adopted from philosophy in around 1980 to describe logical theories developed to support work in artificial intelligence – for instance logical theories capturing the knowledge of the world of a robot. Since then the term has come to be widely used in information-driven science, journalism, industry, defense, intelligence, and government to describe computer artifacts created to support data retrieval, integration, reasoning and an expanding range of further goals.

An ontology is at its core a collection of general terms organized into a taxonomic hierarchy and associated with formal definitions and axioms. A simple example of how an ontology is used occurs where we have a number of heterogeneous data sources, each using different codes or labels to describe the same entities in reality – for instance proteins or weapons – and where we need to unify all the information in these different sources that pertains to each type of represented entity. Ontologies address this problem by providing a data source-neutral term for each type of entity which is then used to tag corresponding data entries in a consistent way. This enables enhanced retrieval and aggregation of the data. It also allows reasoning across aggregated data through appeal to the logical structure of the definitions and axioms of the corresponding ontology terms.

Such strategies have proved most successful where multiple ontologies covering different domains of entities are built in coordinated fashion to ensure interoperability and avoid overlap and redundancy. One method to achieve such coordination involves the application of a hub-and-spokes strategy, resting on a highly general domain-neutral hub, together with successive layers of spokes comprising domain ontologies built out of terms defined as specializations of terms contained in the hub. A top level ontology is an ontology designed to serve as hub in such an architecture.

The objective of the Top Level Ontologies (TLO) project is the standardization of the axiomatizations of current top level ontologies.