SWRL

What types of formal knowledge are needed to support the online pursuit of novel scientific hypotheses or even entire research programs? The past decade in biomedical research has witnessed the advent of huge data sets, large collaborative projects, and the need for both people and computers to make sense of massive quantities of information. Terminologies and ontologies, which permit standardized labeling and structuring of domain concepts, are being developed and used by many scientific communities to support sharing, integration, and management of diverse, distributed data sets. As critical as terminologies and ontologies have become, they are limited in their abilities to model complex definitions of concepts and to allow deductive inference.

Our research group has, in response, created standards-based software methods to edit, manage, and reason with declarative rules, specified in SWRL (Semantic Web Rule Language). SWRL is an extension of OWL (Web Ontology Language), the widely used W3C knowledge representation standard, and is based on a Horn-clause like logic that incorporates built-in operators. Based on the needs of biomedical users, we have extended the set of built-ins to handle temporal representation and reasoning, set-based operations, and ontology-based querying. Our methods are currently available as SWRLTab and Axiomé, two plug-ins developed for the Protégé tool—the most widely used, freely available, open-source program for creating ontologies. SWRLTab provides an editor to author and modify SWRL rules, whereas Axiomé provides a number of techniques for managing SWRL rule sets.

We are currently working on an integrated toolkit for SWRL that is compatible with the next generation of Protégé, supports reasoning over large-scale data sets, and has enhanced features for authoring, managing and executing SWRL rules. We are extending our current methods to provide a full SWRL API for system developers. We are also developing rule-elicitation methods that can guide knowledge acquisition by domain experts not familiar with ontologies, rules or logic. The results of our NIH-funded efforts provide a much needed open-source semantic technology that can encode and manage axiomatic knowledge as rules and that can aid in the querying, integration, and interpretation of biomedical data within ontology-based systems.