- Why is'nt PC target of "semantising movements"? It is because structure of our personal data is fully semantic for us. In other words all information on our computer which we stored, is absolute. Opposite to it: most data on web are of relative, paradigmatic and perspective character. Someone created some information, released it and many is refering it, interpreting its character and content, calling it etc; so we finally may not believe author, nor referer, but only Dennet.
- This is completely the reason for which I previously infirmed 1*document-1*rdf method being in focus of web semantic revolution1 or base of our concentration. I think there will be problems around anonymous documents around, but also sometimes you are not the best to describe your own idea, to sort it in right folder nor to know the "main thema" of it. There is for example many people who would call their gallery "Images of greatest genius in painting" but there is much less people who would at least understand your work or even call you genius for it. So my recycle of rdf (tg) as personal method to sort web, which can be shared and is statistically combined to give the meaning of file for "all of us" seems to solve this Dennets paradox with information on information on web.
- The main thing which, in my opinion, is main reasons why internet is no semantic by itself: each individual web is built primary in reckoning to google like services, and in need for sufficient public (not total, just for this or that social group2).
- I am sure, that (and this is most remarkable sentence of this paragraph, what is good to shortly point out before we continue) if tagging3 of single web would be as difficult as making it favourite4, everyone would do it5.
- Try to imagine, you would may use your computer through google and in-google found directory systems, so instead of instant folders etc, you would see only one textbox and help for it6.
- I know, it is only pure raving but imagine, there was only one "Web directory structure"7 or "filesystem", and everyone should find right location for his web or for link to his partial document. I guess it could be more semantic-friendly than the real estate of web today. Consequently our todays "establishing" semantic web is not clean and pure work, but maybe lot of returning, data mining and maybe in front lines trying to teach pigs singing.
- Now some more serious reading8. Yes if TB-L told, that he want to see some semantic in web - he told it - fine - let's do something! But we may not to take it verbatim, as many of us is doing: trying to find sense of any web page, trying to describe it in rdf. Isn't it stupid to make web where only "Maria married Mario" with their photo is shown, photo eventually called photo.jpg, and this huge indeterministic set of dim informations reinterpret as rdf to header of file? Yes it is. In this and surely in many cases creating a necessary rdf header will be nothing more than requack-quack it for intelligent systems9. Anyway it is still programming what is done with rdf, call it as you please. Remember forever, that best information about content is content itself and doing more than moving it at good place with mouse is not target of being semantic web.
- And for title? Imagine, we have already wide implemented tagging group and our internet is now an easy to search place10. What next? Getting things by sense is routine, but what about structure of documents? Wouldn't it be usefull to see in google11 that some files are blogs, some discussions and some left text documents without any structure? For example to check that some url is blog may be bit more difficult than check that file is some journal. I am sure that to check that page is home of some software may take only counting anchors, refered filetypes and looking for special keywords. In concrete, tasks as: image galleries, personal web pages, directory services, music servers, cute free storages of scientific works, e-book (and other special) download pages, etc are easy to recognize web structures. Now is only on us what to do with this flags, some of next services could do automatic directories from google unsorted stuff, some could only extend own stuff with flags shown and advanced filter options, some could apply back reporting of wrong assignment and self improving algorithms. Look, we aren't talking semantic still, and we are talking about future already12.
- If I was younger, I was playing with idea of "Secure mark". I remembered it while putting anchors to this text. Could you imagine that every internet document would contain very unicate symbol, which would served as point to that document or its part. I mean for what we create documents with some urls, which usually after short period lose their targets. Imagine intelligent anchor which can use any copy of document over the web if there is not original available. For information secure is because it must be exactly found in document where it was stored. So secure mark "2" for text file "12313" is not valid. You could simulate for example intelligent anchor to Daniel Dennet's web pages: if you could link google's "I'm feeling lucky" and query "Dennet homepage" to url you could get pretty good result without implementing my horrible marking globally :)
- If we all had Ford companies and Tchibo companies, and every data on our pages were images of our products and prizing, url system and google would suffice to have fully semantic access to web. But since we are creating different forms of nonsense, extra large nothings, and additional minimum of marriage cards, web needs some intelligent method to see over all these nothing telling "Daniel C. Dennett's Home Page" namings of information islands13.
1) Now I add new infirm: I doubt semantic web is revolution of internet, but I hope it could be new adventure in information and surely successful step (not last) in it.
2) E.g. sorting it in Google directory or in Seznam.cz - for "all" czech people can mean for noone else in final.
3) Ok, I just noticed that maybe also in previous article I was still talking about tagging. Yes I take it almost as synonym to rdf. See paragraph 6 for what I think about rdf else. So if I am talking about tagging of single web I mean making it semantic-accessible at least for you.
4) Note: making file favourite with all it causes (foldering, naming) is way of personal semantising of it.
5) If not everyone, at least same probability that you use Internet Explorer, you will try Firefox and let it installed.
6) Even in DOS was "visible" folder structure. Here is nothing, and I feel bound to repeat, that google-like systems are the most widely used to search, select, sort and "store" new informations.
7) I believe that there is many web directory systems. But also I believe every knows only part of to know, and every knows it at least different way, at worse better or worse.
8) I like Nietzsche, so I am in extasy when I can deny someone's theory, or at best direction of bigger formation of people, so excuse my few next words.
9) Since duck is intelligent system, it would be insulted to be described in rdf and mismatched to hygienic products.
10) Because I am fun of remarks. No, serious, only thing we are trying by get web semantic is clearing up the clutter. But yes, still we are creating new dimensions of nature and improving place to live, don't be sad.
11) I suppose, that after evolving some new semantic and wide access to internet documents, some prime still stay for google-like services. From gyroskopic effect at one side, but from functional at the other side: it may still give you full-text search access, which could stay as best tactic to get source codes etc. In other words it may stay as grep is forever in our hearts and needs.
12) Note to google: if you make very very nice icons to filetypes you surely ... ah, pardon, it is microsoft's patent.
13) That primary naming of such personal data islands, or pillars, is extra egoistic, and has nothing to do with content or tack of its creation is evident and fully tollerable. Don't also name your children with any effort to make it more semantic. And for total dudes (my fictive reader TB-L and DC.D musn't read this warning) do not write any rdf to your baby, nor try to imprint it by unending despotic tyrany and sadism to young child's brain!!!