SUO: Re: Model of Activity and Action in SUO ontology

["John F. Sowa" <sowa@bestweb.net>]

Rich,

I agree:

> Looking at present state of linguistics, it seems
> that the most basic ontology shared by all languages
> is the simple notion of an activity.  Every verb
> in English (see EVCA, for example) takes a set of
> parameters based on its semantic properties.  Lets use
> that verb behavior to convey the notion of activity.
> Then let any single sentence using a verb be an
> instance of the verb's activity model, lets call
> that an action.  

I would generalize it, however, to say "process"
rather than "activity" -- an activity or an action
is a process with an agent.  Some processes, such
as wind, rain, and weather don't have agents.

> Is there anything in CGs, WordNet, SUO, OpenCyc,
> ... that corresponds to this very simple concept
> of activity and actions on objects?

For categories of processes (without addressing the
agents), see

    http://www.jfsowa.com/ontology/process.htm

For physical interactions of processes and causality,

    http://www.jfsowa.com/ontology/causal.htm

For agents,

    http://www.jfsowa.com/ontology/agents.htm

And for the linguistic representations, I recommend
the work on FrameNet:

    http://www.icsi.berkeley.edu/~framenet/
    FrameNet II Home

The frames of FrameNet are very similar to what I have
been calling "canonical graphs", but the FrameNet group
(led by Chuck Fillmore of case grammar fame) has done
a lot more work in specifying and classifying the frames.
Any or all of them could be mapped to CGs.

I just returned from a two-day workshop on FrameNet,
and I'll write a report about my observations, which
I'll post to the mailing list -- probably next week.

John