Property:Text
Jump to navigation
Jump to search
"Text" is a predefined property that represents text of arbitrary length and is provided by Semantic MediaWiki. This property is pre-deployed (also known as special property) and comes with additional administrative privileges but can be used just like any other user-defined property.
For longer texts.
R
You wonder "what do people think of the whole Semantic Web idea?". Well...
I suffered through the 80's Knowledge Representation fad, both academically in the AI program at Edinburgh and as a practitioner at the only company ever to produce a commercial system written in Prolog (that wasn't a Prolog development system.) So I'm familiar with the problems that the Semantic Web effort is attempting to address. Having slogged through real-life efforts to encode substantial amounts of knowledge, I find some of the misty-eyed musings that surround the Semantic Web effort depressing. That "most information on the Web is designed for human consumption" is seen as an obstacle surmountable via tools like RDF is especially sad. On the other hand, I'm always happy to make use of the cool tools that these sorts of things seem to throw off. There's probably a certain Proverbs 26:11 aspect to it as well. +
wahnsinn, sieht escht aus, als würde er aus dem wasser heraus kommen.
was mir einfällt:
eine frau geht nach irland zum studieren. den ersten monat regnet es nur. den zweiten auch. ihre laune sinkt und sinkt... als sie auf die strasse geht, nachdem sie drei monate nur regen mitbekommen hat, schreit sie irgendwann wutentbrannt den nächsten passanten an:
"Regnet es eigentlich immer???!"
Der Passant, der sich angesprochen fühlt, zuckt nur die schultern und meint: "ich weiss nicht, ich bin erst 26!" +
*räusper*
bei dir hat sich irgendwas verschoben.
*guckt auf jahreslinkliste* +
Hm? Was denn? Nur das Stylefile fehlt, oder?
Kann ich nur von daheim aus lösen, habe hier das Stylefile nicht gespeichert. +
den muss ich mir auch noch geben. will aber den ersten nochma davor ansehen.
deeeeeeeeeennyyyyyyyy.........??? magst da nochma rein??? +
Geschichtenerzählen 2.0 ... toller Vergleich ^^ +
:D gefällt mir.
Foobar 2.0 klingt auch viel besser als Mitmach-foobar ;) +
S
In fact, I am in the conference also!! I got to your blog thanks to technorati and it was really nice to see other people's blog on the subject.
It is even more funny to think that you could be right now at 2meters of me and we don't know each other ;-)
Greetings from Heraklion!
FeR +
Yes, that's funny. Maybe we will still meet! :) Any presentations you have? Tried to look in your blog, but my Spanish sucks.
Greeting to Heraklion, from Heraklion! +
How many of the people at SSSW are going to be at the Protege conference next week? +
T
*
+
U
Hi Denny! Nice little essay. I aggree that the unique name assumption would not be a good choice for the semantic web, since we would all have to agree on a common vocabulary for everything before we could even start putting semantic information on the web.
However, I do not understand why inverse functional properties would break with the unique name assumption. In foaf persons are identified with the hash value of their email addresses. With the unique name assumption, we would not be allowed to use two different URIs for the same person. Hence we would not need inverse functional properties and the problem would not occur in the first place.
Greetings,
Ben +
V
W
Hart.
Hast Du das selbst geschrieben? Wow. +
Ja. Wobei der erste Teil erlebt ist. Der Rest ist -- zum Glück -- frei erfunden. +
Harry Potter is an excellent example of a good ontology. +
Uhm, I really would need a bit more of an explanation for that :) +
1) Harry Potter are supposed to be only seven books in the end, and
2) even when it would be wrong.
...
Ok, I shut up. +
You have a good point here, but your examples fail to make it clear.
As you say yourself, foaf:interest is defined to point to a ''web page about the interest'', not to the interest iteself. This means the problem you're trying to show doesn't exist in the case of foaf:interest, and it makes your last two examples bogus. You should have used foaf:topic_interest instead.
The problem, though, is real for many cases. +
Hi Richard,
you are totally right - the example fails. But I took it on purpose, nevertheless: because humans, when they read the predicate "foaf:interest" will think that it means that Subject is interested in the Object.
You are right - and I point it out in the Blog - it actually means the Subject is interested in the Topic of the Object. But I think this is an (intentional) bug in the FOAF-specification. Intentional in order to make writing FOAF-files easier, a bug nevertheless because it creates a cognitive difference between the expected naive and intented specified use of "foaf:interest".
I wanted to stress this cognitive difference with this example. The second reason for selecting this example is to make the reader wary of such subtle differences and mistakes that can happen everywhere. Anything claimed about the Semantic Web should be taken with care, no matter who claims it!
Thanks for your comments,
denny +
I agree, foaf:topic is a poorly defined term. foaf:topic should be named foaf:pageAboutTopic or so. And foaf:topic_interest should be foaf:topic. But that's not what your post talks about, and I'm not sure that mixing this issue into your post is helping to get your point across.
Just a thought. +
Good writeup! A couple of points:
The "overloaded fragment identifier" problem is not really much of a problem. The URI http://semantic.nodix.net/person#Plato can very well be both "a section about Plato in a web page" and "an RDF resource representing Plato in an RDF document". When a web browser asks for the URL, it can be served the web page; when a semantic web agent asks, it can be served the RDF document (through HTTP content negotiation), thus pretty much avoiding the problem.
I'd argue that the anonymous node approach is the right one in many cases. Often, there's no requirement for your stuff being referencable from the outside. In these cases, using anonymous nodes is fine and may save some headaches.
In the long run, maybe there will be sites publishing directories with RDF information about movies or philosophers. Maybe there will be a search engine that lets you search for URIs representing those concepts. Then you can simply use that URI, and don't have to make up your own.
You're mixing up foaf:interest and foaf:topic_interest. The former simply doesn't have the meaning you assume in your examples. Arguably, the FOAF people should have defined them differently, but there's nothing we can do about this now.
I'm looking forward to the last two parts of your series. +
Hi Richard,
Thank you a lot for making the first comment to Semantic Nodix! :)
About your thoughtful points:
I don't like to use content negotiation in order to get the right resource (or rather, representation of a resource, which itself is a resource again). I prefer getting explicitly the resource I asked for, and I feel like that's why URLs are there for: to locate one specific resource. That's why I try to avoid that way.
Anonymous nodes. My only reason against them is that they are not referenceable. You say, most of the time that's ok - I'd claim, well, most of the time you can't know if maybe someone will want to reference it in the future. Why making it impossible? Actually making URIs is very cheap (well, making good URIs isn't, but that's the point of this blogging series), so why not give a reference to every node? The web grew the way it did because we did not claim beforehand to know which resources need to be adressible and which not. It would be a totally different web today if the majority of resource out there were not adressable.
Your last point was answered in the blog and comments of the previous part.
Thanks on your comments,
denny +
A great series of posts. I've always used URIs for all my resources but was begning to question my judgement after seeing widespread use of rdf:nodeID in foaf.
One of the potential applications for foaf, in my opinion, is aggregation and linking across multiple models. I've been scratching my head on how they plan on avoiding nodeID collisions?
Also, have you looked at the Joseki project? What are your thoughts on the RDFWeb API (especially data objects) described there? +
Thanks.
I think HP is doing a great job with Joseki - I mean, I didn't try it myself yet, but the idea behind is the right one. Providing an RDF-WebServer will become essential for the Semantic Web. For me it's much cleaner than the URIQA/MGET-Idea, but according to the mass of discussion ond RDF-IG and now SWIG, I guess, this isn't yet decied... :) +
angeber.
du hast luxemburg vergessen. +
nix angeber! Du hast ja noch afrikanische Staaten dabei, und was weiß ich was :) Außerdem war ich im Gegensatz zu Dir noch gar nicht in Luxemburg! :? +
die afrikanischen staaten machen aber leider meine europakarte nicht schön rot. und ok, dann warst eben nicht in luxemburg - aber ich! *ääätsch*
*g* +
X
Z
hui! danke :o)
hab ich gar nicht gemerkt... +
grz, Tea :o) +
scheintz, als ob mich jemand kennt, auf den ich grad nicht komme? *neugier* +
Ne oder? ^^
Philipp hier. :o) +
oha, DU bist es! :oP
wohl eher wih-tea-went. ;o) +
