SWRL is an expressive OWL-based rule language that can be used to increase the amount of knowledge encoded in OWL ontologies. While semantically a SWRL rule can be considered as an additional type of OWL axiom, the authoring and management of SWRL rule bases requires specialized tools that are not typically present in standard OWL development environments. In this paper, we describe such a tool-called the SWRLAPI-that provides a rich development environment for working with SWRL rules. The SWRLAPI is built on the widely-used Protégé-OWL ontology toolkit. It provides both a set of highly interactive user interfaces for working with rules and a set of lowlevel APIs for embedding rules in knowledge-driven applications. This SWRLAPI has been used to develop a number of technologies to support data integration on the Semantic Web, including a SWRL-based query language called SQWRL and a set of tools that support interoperation between OWL and a variety of information formats.
We have developed a number of open-source tools to work with SWRL. One of the
primary results is the SWRLAPI [1], an extension to the Protégé-OWL ontology
development toolkit. The SWRLAPI provides both an authoring environment for
developing rules and a set of application programmer interfaces that support the
building of rule-driven applications. It has several core software components,
including an editor, a rule engine bridge that supports interoperation with a variety of
rule engines and OWL reasoners, a mechanism for defining Java implementations of
user-defined functions that can be used in rules, and an extensive set of libraries that
can be used to dramatically extend the expressive power of SWRL rules.
This core functionality has been used to build two complementary technologies: (
The SWRLAPI has several core software components, including (
SWRL is a rule language, not a query language. However, many ontology-based applications require the ability to extract information from ontologies in addition to reasoning with the information in those ontologies. To support this knowledge extraction, we have developed a query language called SQWRL (Semantic QueryEnhanced Web Rule Language; [4]) that extends SWRL to support querying of OWL ontologies.
SQWRL is implemented as a built-in library using the standard SWRL built-in mechanism. The SQWRL built-in library contains SQL-influenced built-ins that can be used in a rule to construct retrieval specifications for information stored in an OWL ontology. For example, the following SQWRL query retrieves all persons in an ontology whose age is less than 9, together with their ages: Person(?p) ^ hasAge(?p,?a) ^ swrlb:lessThan(?a,9) -> sqwrl:select(?p,?a) Aggregation, ordering, and counting are also supported. For example, a query to return the average age of persons with a known age in an ontology can be written: Person(?p) ^ hasAge(?p, ?age) -> sqwrl:avg(?age) Like SWRL, SQWRL supports the use of OWL class descriptions. For example, to retrieve all classes in an OWL ontology that are associated with an hasChild property with a cardinality greater than or equal to one, we can write: (hasChild >= 1)(?i) -> sqwrl:select(?i) SQWRL has access to all available SWRLAPI built-in libraries. The ability to freely use built-ins in a query provides a means of continuously expanding the power of the query language. Crucially, it provides an interoperation bridge to deal with other knowledge formats. SQWRL itself is formally based on OWL-DL and SQWRL queries can be expressed only in terms of core OWL concepts. However, the built-ins provides the ability to refer to non OWL DL concepts in queries. A TBox built-in library, for example, allows direct querying of OWL classes and properties, something that would not be possible in OWL DL. Similarly, an ABox built-in library can support the direct querying of ABox entities. For example, to list all data valued properties in an ontology and the number of their direct subproperties, we can write: tbox:isDatatypeProperty(?p) ^ tbox:isDirectSubPropertyOf(?sp,?p) -> sqwrl:select(?p) ^ sqwrl:count(?sp) The SWRLAPI provides a graphical interface to execute SQWRL queries. A JDBClike Java interface is also provided to execute SQWRL queries in Java applications. 4
Data integration is central challenge of the Semantic Web. Semantic Web applications
must typically deal with a variety of information formats. While there are many
dimensions to the data integration challenge, two functionalities are crucial: (
One of SWRL’s most useful features is its ability to incorporate user-defined built-in libraries. This extension mechanism provides a very powerful means of expanding SWRL’s expressiveness and of increasing the types of information that can be reasoned with using rules. In particular, this mechanism can be used to tackle the issue of data integration, which is one of the central challenges of the Semantic Web. The ability to meet this challenge requires the development of a variety of mapping technologies to allow interoperation between the various formats that will be encountered when developing Semantic Web applications. The tools outlined in this paper provide a set of basic building blocks that can be used to construct these mapping technologies. 8th, 2008 from [4] SQWRL (2008). SQWRL. Retrieved http://protege.cim3.net/cgi-bin/wiki.pl?SQWRL from 8th, 2008 from