home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
tsql
/
doc
/
tdbglossary.mail
/
000014_gadia@cs.iastate.edu _Wed Sep 15 14:36:27 1993.msg
< prev
next >
Wrap
Text File
|
1996-01-31
|
4KB
|
98 lines
Message-Id: <199309151937.AA22742@optima.CS.Arizona.EDU>
Received: from ren.cs.iastate.edu by optima.CS.Arizona.EDU (5.65c/15) via SMTP
id AA22742; Wed, 15 Sep 1993 12:37:04 MST
Received: by ren.cs.iastate.edu
(1.37.109.6/16.2) id AA00484; Wed, 15 Sep 93 14:36:27 -0500
From: Shashi K. Gadia <gadia@cs.iastate.edu>
Subject: Re: Ahn/Temporal element
To: ahn@cbnmva.att.com
Date: Wed, 15 Sep 93 14:36:27 CDT
Cc: tdbglossary@cs.arizona.edu
In-Reply-To: from "ahn@cbnmva.att.com" at Sep 15, 93 9:53 am
Mailer: Elm [revision: 70.85]
Hi Ilsoo:
> A temporal element in one dimension is
> a union but NOT a set of intervals.
>
> Greetings.
> If a union is a set operator, doesn't a union of intervals
> result in a set (of intervals) ?
No.
> I don't see the significance of the distinction here.
>
This distinction is similar to: "a relation is not a sequence
of tuples rather a set of tuples. This distinction is
extremely important.
A union of intervals is already a reduced object having
a unique representation, where as a set is not.
> The terms "Multi-intervals" or "Interval sets"
> or simply "Intervals" lean towards a set and
> not union.
>
> I don't see how "temporal element" denotes a union better,
> if a distinction between a set and a union is important.
>
I agree that there is nothing inherent in the term
"temporal element" that says it is a union and not
a set of intervals. One could use a term such as
intervalunion which is self explaining. But in science
when a new and seemingly useful object is born, it is customary to
give it a new name. This allows the concept of that
object to twist, turn, toss and change, and most
importantly evolve through generalization.
> The term temporal element is more generic.
> In the bitemporal model, it is union of rectangles.
> Moreover if your bitemporal space does not contain
> notion of future the bitemporal space reduces
> to triangle below the diagonal. And a temporal
> element becomes union of triangles and rectangles.
> (See Section 6.2 on page P-26 of workshop
> proceedings.)
>
> Term "interval" is used in a general sense too.
> For example, Section 3.15 mentions bitemporal intervals.
>
What when it becomes a union of triangles?
> The term temporal element generalizes to spatial
> or spatio-temporal element. A spatio temporal
> element can be very complex.
>
> A temporal element reminds me of a picture element, which is a pixel.
> This unintended similarity may become more significant
> in a spatio-temporal DB which may deal with images.
I never thought of the interesting similarity with pixels.
Even though in the discrete case any region (hence
a spatial element) becomes a union of pixels this is not
my intention. But that is not what is of essenes here.
A spatial elemen is defined as a union of implementor chosen
primitive regions (e.g. half places, circles, polygons).
In summary the distance between the word "interval"
and "temporal element" is intentional. I think of temporal
element as a starting point where temporal databases start.
Intervals are a cosmic diversion, and have little
significance to me. Even though a temporal elemrnt is
defined as a union of intervals, I use it only to
convery what I have in mind. If I could do it by magic
without using intervals I would to it. To me temporal
element (the concept) is the most primitive notion in
temporal databases.
>
> Regards
> Ilsoo Ahn
>