[sc34wg3] Topic Map View
Patrick Durusau
sc34wg3@isotopicmaps.org
Fri, 10 Sep 2004 22:44:41 -0400
Dear Lutz,
By way of explanation, Lutz sent this post offlist but I obtained his
permission to reply both to him and the SC34 list.
Dipl.-Wirtsch.-Inf. Lutz Maicher [Universität Leipzig] wrote:
> Greetings,
>
> some weeks ago the term "Topic Map View" was introduced by Patrick
> Durusau.
>
>
>>For those of you who were unable to attend the Montreal reference
>>model workshop meeting, be aware that a new term has sprung into >>being!
>>
>>"Topic map view"
>
>
> Unfortunatley the discussion about Topic Map Views stopped before I
>was really confident with the new term.
>
>>From my viewpoint a Topic Map View is a "visualisation" of a
> "materialisation" of a Topic Map in a given Application. If this
>assumption is true, I will interpret the Opera-Topic-Map showed in the
>Omnigator as a View of this Topic Map. This view consists of the
>opera.xtm and some rules (coded in the Omnigator) how this Topic Map
>Syntax can be transformed to the web pages of the Omnigator.
>
Let me defer answering this question until we have examined your main
questions.
> My main questions are: Which rules for interpreting the Topic Map are
>really part of the Topic Map View?
> Why does we need equality and merging rules within the Topic Map View?
>
> Perhaps, these questions should be discussed by using an examples.
>
>
>>The answer that makes the topic maps paradigm scale to such uses, is
>>that of the "topic map view."
>>
Note that in Newcomb's formulation, a 'subject-centric information
environment' = a subject proxy.
>>Newcomb suggested the following formulation of topic map view:
>>
>>**Newcomb**
>>(a representation of a topic map of type X)
>>+ (a set of rules for interpreting topic maps of type X)
>>--------------------------------------------------------
>>= (a topic map view) -- an interpretation in which everything has
>> become explicit as a set of subject-centric
>> information environments.
>>
>>In other words, a "topic map view" is a particular way of seeing some
>>particular data as a particular topic map.
>>
>>So what's the exact nature of a "topic map view"? What distinguishes
>>a "topic map view" from anything else?
>>
>> (1) In a topic map view, you can always detect when two
>> subject-centric information environments have the same
>> subject, and
>>
>> (2) When subject-sameness has been detected, the two information
>> environments can be merged, becoming a single,
>> more-comprehensive information environment, with the same
>> subject as its invisible heart.
>>**/end Newcomb**
To start the conversation off at a common starting point, let's start
with more familiar examples of maps, before addressing the rules that
are required for a topic map view.
1. Point of view
Take a street map and a tourist map of a locale that is of interest to
you. Both maps are assumed for the following to be of the same
locale.
Comparing the two side by side, you will discover, in most cases, that
the two maps differ in some respects, despite their being maps of the
same locale. One may show more streets than the other or items such as
schools or hospitals, these being more likely on the street map. The
tourist map, on the other hand, may have special symbols for museums,
historical sites or other items thought to be of interest to
tourists. Both will probably show public transportation systems, major
street intersections and bridges.
Both maps are maps of the same locale but are immediately
recognizeable as different 'views' of the same locale. Neither is more
'correct' or 'right' than the other, each being suitable or thought to
be suitable by its author or its users, for a particular purpose.
The authors of these different maps chose the features of the locale
in question that they would represent in their maps. A street map will
not (usually) represent topographical features and a topographical map
will not represent political divisions, all of which are 'present' for
any given locale, depending upon one's point of view.
So, the first choice that a map author makes is what will appear in
the map, out of a potentially unlimited number of choices, and those
choices in sum are the 'point of view' of a particular map.
2. Features on a map
Another feature of the maps in question is that each map has a unique
location for everything that is present on the map. It would be a very
odd map that had multiple locations for a single subway station or a
particular museum building.
If the map lists subway stations, it may also lists the trains that
stop at each one, but there is a difference between the listing of the
trains and the subway stations.
The subway stations each have a unique location on the map, while the
trains may or may not appear at any particular subway station.
[Digression on subject proxies vs. properties of subject proxies:
Note that in the foregoing example, subway stations are subject
proxies and the trains that stop at them are properties of subject
proxies. The difference, even for a topic map is subtle but profound.
Taking the tourist map in front of you, prepare a note for each of the
sites you have visited, using variant forms of your personal name. You
can list your full name, only your family name, a nickname, etc. Do
the same for all of your family members and several friends.
If you look at the the map for all the visitors to a particular site,
you will find the list of all the people you have listed.
If one were to look at the the map for all the places any individual
has visited, you would have to examine each location. Why? Well there
is no subject proxy for the person in question with a subject identity
property to insure that all the information about that person
coalesces to a single point on the map.
A subject identity property could be a PSI or some combination of
properties that in the view of the map's author, uniquely identify you
for the purposes of coalescing information about you.
Two essential points here:
1. The author of the map determines what will be represented in his
map (can be changed by authoring another map by any means but that
map must adhere to all the rules established by the other map's
author)
2. The author of the map determines on what basis comparisons are made
for the purpose of determining if two pieces of information should
be recorded at a single location on the map.
Note: Some bases for comparison will work better than others, varying
depending upon the situation.
End digression]
3. Following the rules of the map:
Let's imagine that we have a map of cell phone tower locations that is
generated from the data held by the cell phone company. That map,
displayed on a computer screen, displays the physical location of the
cell phone tower, the 'best' and alternative routes for servicing.
From the data held by the cell phone company , we can produce a map
that has icons (subject proxies) for all the users of that tower
beyond a certain number of minutes. Certainly a legitimate map, but
probably not as useful as the first one for the technician who is
going to fix a faulty cell phone tower transmitter.
This is another example of a 'view' of data but I offer it to
illustrate that the integrity of the map depends upon the rules under
which it was constructed being part of the basis for any view of it.
Granted, one can always construct another map, but that has no impact
upon the integrity of the original map.
Using a map to view data and not using the rules under which it was
constructed is either:
1. Misuse of the map, possibly leading to incorrect conclusions, or
2. Authoring of another map for viewing the data
which are distinct cases.
Apologies for the long lead in but I think it will provide a context
for my answers to your questions:
> Which rules for interpreting the Topic Map are really
> part of the Topic Map View?
All of them.
Recall that the street and tourist maps were useful only because:
1. Every thing in the map has a single location
2. A distinction is made between things with a single location (and an
implied subject identity property, though tourist map makers are
unlikely to use that term) and properties of those things.
Recall that topic maps make explicit the process that the maker of a
physical map does mentally by placing something at a particular
location and attaching properties to it.
With computer systems, we don't have to look and see if we have
written down something before, so long as we have rules the computer
can follow in detecting duplicates and creating a single location no
matter how many times we have entered information about the same
thing.
Since the information we are recording for a single thing may be a
list, such as visitors, trains that stop at a location, etc., we must
have rules for how that information gets combined at a single
location.
> Why does we need equality and merging rules within the Topic Map View?
>
The topic map view produces a map of the data according to the rules
of the topic map. Without equality or merging rules, we would be
unable to:
1. Have a single location for all the information related to something
in our map, and
2. We would not have any rules for combining the information we gather
to a single location.
>>From my viewpoint a Topic Map View is a "visualisation" of a
> "materialisation" of a Topic Map in a given Application. If this
>assumption is true, I will interpret the Opera-Topic-Map showed in the
>Omnigator as a View of this Topic Map. This view consists of the
>opera.xtm and some rules (coded in the Omnigator) how this Topic Map
>Syntax can be transformed to the web pages of the Omnigator.
>
Finally, your first question last! (sorry)
It is possible to have a topic map view of a topic map (a document
that purportedly represents a topic map, perhaps notated in XTM
syntax). A topic map (in that sense) is, afterall, data and subject to
the same operations as any other data.
However, a 'topic map view' is an abstract thing. And a 'topic map
view' can be rendered in many different ways. It is important to
distinguish between a 'topic map view' and a rendition of such a view.
(Noting that in casual conversation we may say 'view' when actually we
mean a 'rendition' of a particular 'view' for display purposes.)
I will be returning to Atlanta tomorrow from San Jose so I will be
offline for most of the day. Looking forward to further discussion.
Hope you are having a great day!
Patrick
--
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
Patrick.Durusau@sbl-site.org
Chair, V1 - Text Processing: Office and Publishing Systems Interface
Co-Editor, ISO 13250, Topic Maps -- Reference Model
Topic Maps: Human, not artificial, intelligence at work!