Raytheon BBN Technologies

Semantic technologies revolutionize the way in which we address access to and integration of large, diverse data resources. However, to take advantage of these new capabilities we need to deal with the issue of ontological alignment. People observe the world from differing perspectives and semantic technologies allow us to capture these perspectives in a logical and meaningful manner that encourages more effective sharing. To an auto manufacturer, a truck is a product. To a traffic analyst, a truck is simply a vehicle to be counted. To a shipping company, a truck represents a unit of transport capacity. In order for these assumptions about the world to be shared and reused, conceptual alignment is required.

Aligning conceptual models is a significant challenge. Each model may use different terminology and different levels of detail to formalize ideas and the relationships that define them. Current solutions often require countless hours of costly manual intervention, are complicated to implement, and can be very rigid (brittle). Raytheon BBN Technologies solves this problem with well-developed ontology mapping tools, such as “Snoggle.” Snoggle is a Raytheon BBN-led open source tool that greatly simplifies the process of ontological alignment. A key feature of this tool is its intuitive interface that makes data model mapping quick and easy, while preventing errors associated with manually editing complex rules. Snoggle exploits our innate ability to think visually to simplify the alignment process.

Snoggle is a graphical, ontology mapper based on the Semantic Web Rule Language (SWRL) and assists in aligning ontologies represented in the Web Ontology Language. It allows users to visualize ontologies and then draw mappings between them on an intuitive graphical canvas. Users draw mappings as appropriate, and the tool then transforms them into SWRL/RDF or SWRL/XML for use in a knowledge base.

This intuitive graphic process replaces the laborious and error-prone task of manually writing rules in a text editor. The graphical canvas in Snoggle is divided into two regions: “from” and “to.” The "from" region contains a structure in the source ontology, and the "to" region contains the corresponding structure in the destination ontology. This arrangement models a user’s common view of mapping structures. Arrows are drawn between the two regions to represent mappings. For example, a Person who is 6 feet, 2 inches tall in one ontology, may be represented as an Employee who is 74 inches tall in another ontology.

Snoggle also supports the default SWRL built-ins and allows for easy insertion of your own custom SWRL built-in definitions. Built-ins can be thought of as functions that perform calculations or actions as part of a rule. An example built-in might perform mathematical operations (for unit conversion, for example) or transform a data from one type to another. Consider, for example, a mapping between two ontologies, where one uses a numeric representation for date (e.g. 20070315) and the other users a string (e.g. March 15, 2007). A custom built-in could be used to convert between the two representations. Snoggle’s ability to provide built-in customization opens up entirely new possibilities for handling and easing alignment, all without the need for another tool or without sacrificing the use of semantic web standards.

Complex rules are easily created by dropping types defined in the ontologies onto the canvas and linking them with relationships. Snoggle also provides the ability to export these mapping rules for use in a knowledge base capable of SWRL inference.