[Gfo-users] Re: Time and GFO

[Frank Loebe]

Hi Veronika,

so far we have only initial conceptions on extending GFO-time with
calendar-like notions (e.g., time units). Unfortunately, there is nothing
available in written form yet. I'll provide some remarks on the conceptual
connections with W3C/OWL-Time. I cannot say anything about the
interrelations of the axiomatizations yet, apart from having noted some
differences which will have to be accounted for formally, also wrt the OWL
formalizations available. But in principle and conceptually speaking, I
think the two time theories are compatible.

I see the following correspondences as being conceptually closest to notions
related to GFO-Time entities:

w3c:TemporalEntity & gfo:Temporal_region
w3c:Instant        & (a pair of coincident gfo:Time_boundary)
w3c:Interval       & (requires longer discussion)
w3c:ProperInterval & gfo:Chronoid

With respect to instants, I propose to view them as pairs of coincident time
boundaries and think of the meaning of affected properties with appropriate
adjustments. E.g., if w3c:hasBeginning links to a w3c:instant I which is
understood as a pair of coincident boundaries (b1,b2), this should refer to
a gfo:has_left_time_boundary link to b2, whereas w3c:hasEnd I transforms
into gfo:has_right_time_boundary to b1. (The order in the pair is derived
from the temporal order of two meeting intervals, one ending in b1, the
other starting in b2.)

The remaining notions require further analysis wrt GFO and will involve
other notions apart from only temporal ones (esp. symbols and
representation):
w3c:DateTimeInterval 
w3c:DurationDescription 
w3c:DateTimeDescription
w3c:TemporalUnit
w3c:DayOfWeek

Most considerations in W3C-Time appear relevant and appropriate, yet some of
them would be treated differently wrt some details in GFO, and in some cases
I'd expect that further analyses and some extensions are required. Sorry for
not being clearer in these respects today. Altogether, I think you can use
these notions conceptually, because they must be added to GFO-Time anyway in
the future. Thus, a question for me is, wether you use W3C- and GFO-Time
only conceptually, or you plan to employ the OWL versions together? How does
OWL relate to the implementation of your clinical database? This affects the
above answers or demands additions to them.

If you have further questions, please let us know.  

Best regards,
Frank

> Hi, 
> 
> we are developing a clinical database to describe clinical 
> research data. Therefore we have to talk about time, for 
> example the exact timepoint of a measurement, how long a 
> surgery takes etc. We're using the GFO to model events in our 
> database, but we also wish to describe these events using the 
> W3C time ontology so that we can talk about minutes, hours, 
> days, rather than time-boundaries.
> Do you have any "paradigm" or best-practices for doing this 
> in GFO instances?
> 
> Thanks!
> Veronika