Knowledge Representation (KR) is a field in computer science (and AI especially) that aims to find methods to model real-world information in ways that can be understood, and reasoned about, through logic and computer programming. Many different representations are used, such as terminologies and taxonomies. In particular, ontologies are specific formal representations of knowledge, modelled as a set of concepts and the relationships between them.
Numerous ontology languages exist, and are generally well supported, but many ontology creation programs have a problem that it is difficult to create the representation "on the fly"—knowledge of the planned structure of the ontology is required before ontology editing can start.
SKOT is a tool for this initial sketching stage, where lists of words (representing concepts) can be input, then placed onto a diagram as blobs, and freely arranged and linked together as appropriate. Once the user is done, they can export the resulting file into an OWL ontology file so that it can be edited further.
Sources are available through Subversion here, and a built version (as a zipped .jar file with libraries) is available on the sourceforge download page. SKOT development is currently on hold due to the end of the associated project, but if you have specific feature requests or bugs that need fixing then I'll see what I can do about it.
Mark can be contacted through the sourceforge support page or by email at this email address.