Pr4 UG mi Selection
Selection
comment Editing Glossary Outline
| Protege 4 User's Guide |
|---|
| Miscellany |
| Colophon |
| Conventions |
| Editing |
| Feedback |
| Glossary |
| Introduction |
| Outline |
| Page List |
| Page Report |
| Resources |
| Selection |
| Startup |
It's quite possible that Protege 4 may not be the best tool for your needs. There are other versions of Protege and, indeed, other ontology editors and tools for knowledge management. Each of these tools will have benefits and drawbacks for a particular use.
By pondering some of the questions below, you may determine that Protege 4 is not what you need. Conversely, you may come away with a better appreciation of why Protege 4 is just the tool you need. In either case, it will be time well spent...
Is an ontology really what I need?
Ontologies are necessary for mechanized reasoning and useful for other purposes such as database design and creation of standard lexicons. The act of creating an ontology can be very valuable; it is almost certain to uncover ambiguities and other kinds of confusion.
However, ontologies can also be difficult and time-consuming to create. If all you need is a glossary or a thesaurus for human use, creating an ontology is not necessarily the best approach.
What tools do I need?
Simple ontologies can be created using general-purpose tools (eg, outliners, spreadsheets, text editors), used in conjunction with stand-alone (eg, command-line) tools.
However, as an ontology grows in size and complexity, this approach may become unwieldy. Also, an integrated environment can provide convenient access to specialized tools, views, etc. Consequently, a variety of ontology editors have been developed.
Traditional ontologies are composed of "frames" (ie, records), each of which has some number of attributes. Protege 3 supports this approach, but Protege 4 does not. Instead, it assists in the creation of OWL-based ontologies, which are built out of sets of (asserted and inferred) binary relations.
If what you want is an OWL-based ontology, either Protege 3 or Protege 4 may work for you. However, Protege 3 only supports OWL 1; if you need OWL 2 (and beyond), Protege 4 is the right answer.
If your ontology will be based on something other than frames or OWL, you should look elsewhere for an appropriate tool. For example, if you expect to be creating your ontology using RDF and RDFS, you may wish to investigate tools such as TopBraid Composer. See ontology editor (WP) for more examples.
Do I want to be an early adopter?
Protege 3 has been around for decades. It is stable, robust, well-understood and -debugged, etc. It also has a plethora of useful plugins and a large user community. Protege 4, in contrast, is relatively new software, so some of these benefits may be diminished.
If you want to be on the leading edge of OWL ontologies, Protege 4 is clearly the right choice. If, however, you simply want to create a robust OWL 1 ontology, Protege 3 may be a better choice.
Fortunately, if all you need is OWL 1, you can use either version: OWL 1 ontologies can be freely exchanged between Protege 3 and Protege 4. Indeed, some users perform this action routinely, in order to take advantage of specific program features.
Will this be a collaborative effort?
Protege 3 and Protege 4 are desktop applications. If your effort will be collaborative, you may want to investigate Collaborative Protege and/or WebProtege.
