gist: the minimalist upper ontology

What is gist?
gist is a an Upper Ontology. It is different from other upper ontologies in that we have attempted to do two things simultaneously:

- cover a very broad range of future applications

- cover them with the fewest number of concepts

Why isn't it smaller?

One of the first questions always seems to be some variation on: "why didn't you go all the way up the abstraction tree to entity/relation/attribute, or class/subclass or thing?" It is because those aren't really semantic concepts, or at least they are not concepts in any domain. They are concepts about the structure of the information. You can put most concepts in any of those structures.

Getting Started
gist is freely available. Click the download on the left to get either a current version of gist in owl, or the presentation that describes the content. More white papers, etc. will be posted soon.

Warning on versioning

I haven't set up version control yet, so please use this on a experimental basis for the time being.  We will be setting up proper versioning, (and an easier to type url!) but for now just take a copy an use it as it.

Coming Soon

I intend to start a dialogue around the use of this. There will be a wiki, some sort of message board and/or a blog. In the meantime, if you'd like to stay apprised, or if you'd like to help out, feel free to send me an email at:

mccomb(at)semanticarts.com

(replace the (at) with  @ as I'm trying to foil fhe spam harvesters)

On the left are three versions of the same diagram (visio, svg and jpg).  They aren't quite right, and I'm on the road right now, but I know there are a few of you who may be trying to use the ontology and this may help.

The main purpose of the diagram is to give an idea of the highest level concepts, and in particular which ones are disjoint from which others. 

The ones at the very top have some overlaps that I haven't gotten around to laying out how they relate to the rest of it, and there are a few classes missing, but this should be a better place to start.

Regarding the class names. 

Keep in mind, the class names are just "labels."  I've tried to come up with conceptual definitions, and then use a lable that would be evocative of that concept, once you know what it meant.

I mostly tried to avoid more precise lables (erdurants and perdurants for instance) because they become a bit unweildy to newcomers.  I have capitulated in a few cases (extensional and intensional, and durableTemporalRelations) largely because I couldn't think of a more common term that wasn't hopelessly overloaded. 

There are a few areas where this approach may cause some confusion.  For instance: "living thing" is not meant to be a definition of "living things" the set, as so composed, are things that have been born, and are individually identifiable.  So it includes most mammals, to the extent they are identified in our systems (a vetrinarian system for instance).  Because of open world reasoning, it is really "things that were alive at some time" and includes things that are dead, or might be dead (me for instance, however "I'm not dead.  I'm actually feeling much better.") I haven't been precise yet, as to whether germinating constitutes being born, but I'm thinking it does, which suggests I need to allow landmarks to overlap with living things (trees)).