2.2 The Semantic Web
2.2.5 Ontologies
2.2.5.6 Ontology Devlopment Tools
A number of development and editing tools are available to ease the complex and time consuming task of building ontologies. Tools such as Kaon30, OileEd31, and Protégé32
provide interfaces that help users carry out some of the main activities required for developing an ontology. Selecting the most appropriate editor, however, is a challenge because each ontology construction initiative requires its own budget, time, and resources. To help overcome this challenge, Singh & Murshed (2005) proposed criteria to evaluate ontology construction tools. The criteria include functionality, reusability, data storage, complexity, association, scalability, resilience, reliability, robustness, learn- ability, availability, efficiency, and visibility. Protégé and OntoEdit33 Free (the
predecessor to Ontostudio), were evaluated by Singh & Murshed (op. cit) using this criterion. The evaluation concluded that the editors provide similar functionality.
30 Kaon version 1.2.7 available at: http://kaon.semanticWeb.org/ 31 OildEd version 3.5 available at: http://oiled.man.ac.uk/ 32 Protégé 3.2 available at: http://protege.stanford.edu/ 33 http://www.ontoknowledge.org/tools/ontoedit.shtml
A survey of ontology editors conducted by Denny (2002) classified available commercial products as either standalone editors designed exclusively for building ontologies in any domain, or editors that are part of commercial software suites designed to deliver broad enterprise integration solutions. Denny (op. cit) concluded that non-commercial editing software were generally the outcome of academic and government funded projects investigating the technical application of ontologies, with some intended for building ontologies in a specific domain. The later type of editors were still capable of general- purpose ontology building regardless of content focus.
Probably, the most comprehensive survey of ontology editors conducted to date is that of Damjanoviæ et al. (2004). Their survey was based on the following six criteria: 1) general description of the tools (such as information about developers, releases and availability); 2) software architecture and tool evolution; 3) interoperability with other ontology development tools and languages; 4) knowledge representation paradigm (knowledge model used); 5) inference services attached to the tool; and 6) tool usability. Damjanoviæ et al. (op. cit) concluded that:
• Ontology languages from the pre-XML era have matured. Unfortunately,
Damjanoviæ et al. (op. cit) found that unlike these, tools and ontology development languages from the XML-era still aren’t mature. Hence, the tools are continuously evolving. New research areas emerge from deploying intelligent Web services (a combination of the emerging Semantic Web and Web services technologies), but require new research efforts, new development tools, and new tools for dynamic management of the Web.
• From the criteria for ontology development tool extensibility, Damjanoviæ et al. (op. cit) reported a trend of further adaptation of existing ontology development tools to the new Web standards (W3C recommendations), such as RDF (Resource Definition Framework), OWL (Web Ontology Language). They also stressed the importance of the newly proposed ISO standard, known as CL (Common Logic) that will be compatible with all the accepted W3C standards. Damjanoviæ et al. (op. cit) state, however, that this trend is not equally represented in all of these tools. This is because certain problems relate to ontology development tool interoperability. Usually, different research groups develop different tools and as a consequence, ontology development environments and tools are not interoperable. These tools have different
knowledge models, use different technology, and it is often difficult to integrate them. More recent ontology development tools allow for exporting and importing ontologies in XML and other markup languages as a means of exchanging ontologies between the tools. This can improve interoperability.
• Like the ontology development tools extensibility criteria, the portability criteria pertain to the ability of a tool to adapt easily to a new environment. Damjanoviæ et al. (op. cit) say that a good example of this is Protégé, which has a component framework for easily integrating other components via plug-ins. Thus, it was concluded that Protégé brings with it a great potential to expand, and also to adapt itself to the new (ontology-based) development environment. But this is not the case with all tools.
• The ‘ease-of-use’ criteria was claimed to be very important since it implies a necessity to use intuitive screen designs for anyone who will work in the area of ontology development, maintenance, deployment, merging, and update. However, current ontology development tools require their users to be trained in knowledge representation and abstraction.
• Finally, the study found that the use of ontology development tools in the sense of discovery and search criteria is important in the Web environment to find some potentially interesting new knowledge. Moreover, this criterion is related to the ability of validating, evolving, and maintaining this knowledge.
A number of popular ontology editing tools were experimented with while conducting this research in order to gain first hand knowledge of their functionality. Table 2 is a list of the ontology editing tools that were tried by the researcher:
Protégé 3.2, which now supports OWL, is one of the oldest and most widely used ontology editors available today. It allows the user to define and edit ontology classes, properties, relationships and instances using a tree structure. Ontologies can be exported into a variety of formats including RDF(S), and XML Schema. Protégé 3.2 was the most user friendly and functionally superior tool tried. This assessment was based on the fact that the Protégé platform supports two main ways of modelling ontologies via the Protégé-Frames and the Protégé-OWL editor, and the fact that Protégé has many useful plugins. A visualisation tab, for instance, called OWL Viz allows ontologies to be viewed in graph form and exported to a JPEG file. The application also includes a SPARQL query tab.
Developer Product Availability Language
Support
FZI – AIFB
http://kaon.semanticWeb.org/frontpage KAON 1.2.7 Open source KAON RDF(S) IMG (University of Manchester)
http://oiled.man.ac.uk/index.shtml
OilEd 3.5 Open source RDF(S) OIL DAML+OIL OWL SHIQ Ontoprise http://www.ontoprise.de/content/e3/e43/index_eng.html Ontostudio 1.4 Freeware Licenses RDF(S) OWL F-Logic OXML SMI (Stanford University)
http://protege.stanford.edu/ Protégé 3.2 Open source XML RDF(S) XML Schema OWL KMI (Open University)
http://kmi.open.ac.uk/projects/Webonto/ WebOnto 2.3 Free access OCML RDF(S) Mindswap
http://www.mindswap.org/2004/SWOOP/ Swoop 2.3 Open Source RDF(S) OWL