Re: Afghanistan War Diary as topic map in Maiana

Folks, 

The heat and rancor are not going to lead to a happy conclusion for this
thread.

I think Lars meant that holding data in topic map format, without more,
isn't very interesting, and I have to agree with that.

What makes data in a topic map format *interesting* is the ability to
merge it with other data, that identifies some of the same subjects
differently. 

Rather than saying "without an ontology," I would say "without stated
conditions for merging" but that may be quibbling over terminology.

I think a more productive course for this conversation to take would be
to ask what other data could be merged with this data set, on what basis
could/should such a merging occur and if that isn't possible now with
Maiana, what topic map software does support such features? (A
subsidiary topic being when will we see a release of Maiana?)

Hope everyone is having a great day!

Patrick

PS: I don't think that Lars has something against the Topic Maps Lab nor
do I think the Topic Maps Lab has it in for Lars. 

I do think everyone is seizing upon nuances and implications that simply
aren't apparent to most of the rest of us. 

A little more generous reading of emails, usually composed hurriedly,
would go a long way. 

On Fri, 2010-10-29 at 12:15 +1300, Murray Altheim wrote:
> On 29/10/10 11:21, Lars Heuer wrote:
> > Hi Murray,
> >
> > [...]
> >> Why the snarky attitude? All Lutz has done is to inform this community
> >> of an interesting application of Topic Maps. If you don't believe there
> >> is any value, then why tell us you think it's "great"?
> >
> > I don't deny that it would be great if WikiLeaks uses Topic Maps /
> > RDF.
> 
> No, you don't deny it, but you contrarily continue to state you find
> either no or limited value. Others obviously do: in the past few hours
> both myself and Miles have publicly stated we see real value in the
> project.
> 
> And while you say you're not eager to feed trolls you seem to be
> exhibiting troll-like behaviour. Why? Do you find this productive?
> Enjoy a good tussle? You seem to have an agenda. Why not tell us
> rather than play around?
> 
> >>     * At very least this demonstrates a way to map and query a large
> >>       corpus of documents that didn't originate in a form designed to
> >>       be mapped. That is in itself a success, simply as an open example
> >>       of how a translation project can be approached (i.e., a public
> >>       demonstration is always welcome, no matter how simple);
> >
> > Yes. I can also translate WikiLeaks to JSON. Nobody denies that. XML
> > is also translatable to JSON. Is any translation to JSON JSONish? I
> > don't think so.
> >
> > Is any mapping Topic Maps-ish? I don't think so.
> 
> Your question is tautological. A translation of WikiLeaks into JSON is
> JSON-ish, a translation into Topic Maps is Topic Maps-ish. These
> mappings permit the content to be more readily used respectively
> by JSON or Topic Map-based tools, which add value. Demonstrably so.
> 
> > [3 further arguments]
> >
> > I just say: You may be wrong if you translate WikiLeaks to Topic Maps
> > without an ontology. You may translate Atom or MARC to Topic Maps 1:1,
> > but where is the value?
> 
> Please don't lecture me on ontologies, certainly one of the most abused
> terms of the past few decades. Machine reasoning based on this material
> would be problematic at best*. There is no "ontology" required in a
> simple mapping, which is all this purports to be, and indeed, a simple
> mapping without interpretation is at least semantically safe, where what
> you are suggesting is not so. Lutz has already replied and shown that
> what they've done permits further development, and further value.
> 
> >> I don't know what if any backstory there might be in your desire to
> >> be publicly critical, but on the face of it, that you can't see any
> >> value in this suggests either a desire to denigrate others' successes
> >> or a lack of imagination.
> >
> > Being published as Topic Maps (or RDF) 1:1 without semantics/an
> > ontology adds no value at all.
> 
> Nonsense. Publishing WikiLeaks with an "ontology" created by an external
> party (such as you, I, WikiLeaks, or anyone not involved in the origins
> of the content) would be making ontological commitments that are not
> justified from the content or by any authority. It's better to let
> humans (users of an end-user application) make their own inferences
> from the material -- where they are making explicit their own
> ontological commitments -- than to provide an "ontology" that hides
> those commitments and does violence to meaning in inferring an
> unjustified level of truth, confidence, or understanding. I for one do
> not want to be part of promulgating such falsehoods, particularly for
> such politically-charged material.
> 
> In other words, if you want to be "wrong" the best avenue is provide
> an externally-produced ontology.
> 
> Murray
> 
> * as reported in The Guardian yesterday, if all variants of Mohammed
>    were romanised from the Arabic in the same way it would be the most
>    popular name in the UK. Misidentification of people continues to be
>    a major problem in military intelligence and has led to false arrests,
>    imprisonment and torture. This is "being wrong" writ large and painful.
> ...........................................................................
> Murray Altheim <murray10 at altheim dot com>                       ===  = =
> http://www.altheim.com/murray/                                     = =  ===
> SGML Grease Monkey, Banjo Player, Wantanabe Zen Monk               = =  = =
> 
>        Boundless wind and moon - the eye within eyes,
>        Inexhaustible heaven and earth - the light beyond light,
>        The willow dark, the flower bright - ten thousand houses,
>        Knock at any door - there's one who will respond.
>                                        -- The Blue Cliff Record
> _______________________________________________
> topicmapmail mailing list
> topicmapmail-Zo64W7twoUFWk0Htik3J/[email protected]
> http://www.infoloom.com/mailman/listinfo/topicmapmail