How Google Will Have Achieved The Semantic Web

alfaromeo points to a business feature (mysteriously available already) by one Paul Ford called "August 2009: How Google beat Amazon and Ebay to the Semantic Web." So read on for a bit of potential history from five years in the future.

Comments

Slashdot purpose

[someguy456]

So I guess since ./ couldn't handle the past, and is failing miserably with the present, it will now resort to fortune-telling?

Editors, could we at least keep the dupes down? :)

But none of this ever happened

[Halcyon-X]

because all of the patents to do so were tied up between various companies that didn't want to cooperate with each other.

Re:Heh

[primordial ooze]

Remember back when we all thought that XML was going to achieve the semantic web by making good search engines unnecessary?

Not really, and XML is still such a recent development that to say "Remember when" is silly if not outright disengenuous. I was at the SGML '86 conference in Boston where the XML initial draft was presented. That's less than ten years ago. Can you name a information technology that reached anything like its full potential less than a decade after its first mention?

Re:Heh

[jZnat]

Don't quote me on this (seriously, because if you do, I will cut you), but I thought Microsoft was migrating to XML usage for a lot of their proprietary formats finally. I think it's a good idea, but then again, what if they patent their XML formats?

Yeah, just letting you know that XML is actually going somewhere.

Re:Heh

[KefabiMe]

I RTFA. Intriguing, but it would be a huge struggle for Google to become like anything in the article. There's too much money in having the right information at the right time.

"Now XML has gone nowhere except as a set of popular libraries for cross-language data serialization..."

XML is still getting more popular and more accepted with each passing month.

The biggest issues are that there are a few monstrous companies out there that want to control the standard of how information is shared, and mutate XML into some proprietary form that their company can control.

XML is a good thing, like most standards. Standards can fall short at times, especially when the uber-companies start trying to fight for control over them. I believe that this fight for control will do more to prevent the easy transfer of data, more than any problem with XML itself.

BB

[jals]

I started off reading this and gradually got quite excited by the ideas presented.

About half way through I mistakenly thought I was reading an online copy of 1984.

The benifits of this happening sound fantastic. It just sounds very cool for everyone to be connected like that - which is what scares me even more. Here is an absolutely huge privacy concern; and it has me totally excited about the prospect of it happening.

Sorry to go slightly off topic, but it's things like this that worry me a lot, that a possible 1984 scenario could disguise itself so well that even a person like me - who is verging on (if not already there) being a member of the tin foil hat brigade - excited by the very idea of it.

Five years into the future?

[lurker412]

Give me a break. Machine translation of foreign languages has been promised "in the next ten years" for the past 40 years. Unfortunately, the state of the art is still very close to:

"The flesh is willing but the spirit is weak" in English translates to "The meat is full of stars but the vodka is made of pinking shears" or suchlike in Russian.

The semantic web is a wonderful dream, but it is certainly going to take more than five years to become a reality. Like voice recognition, the semantic web requires a solution to the natural language problem to be implemented successfully. Don't hold your breath.

Re:Five years into the future?

[LetterJ]

Sure, colloquial translation is lacking, but text that is deliberately composed to *be translated*, goes through suprisingly well.

I mean, would you use "out of sight, out of mind" in a conversation with someone who had only a couple of years of English classes without having to explain it? Probably not. Rather, you'd most likely use a smaller vocabulary with fewer long phrases and idioms. If you do that with your text intended for translation, it does pretty well.

The goal of most translation is the ability to communicate and things like Babelfish have allowed me to communicate with users of my software from around the world with little difficulty, each using our own language. Are we going to be collaborating on great literature? No. However if I can get Babelfish level translation done portable and cheap, traveling will be much more enjoyable.

Re:Heh

[Anonymous Coward]

Remember back when we all thought that XML was going to achieve the semantic web by making good search engines unnecessary?

Nope. I remember a bunch of people with no clue hyping it up as such, but anybody actually involved with XML in any technical capacity, including the creators, understood that it was simply a standardised syntax for file formats. So-called pundits jumped on each others' bandwagons in touting it as some kind of miracle, but anybody who actually knew what they were talking about wouldn't make claims about XML that you reckon.

Re:The French and the Germans

[stefanlasiewski]

Looking at the resultant cars, who do you think had a better approach?

If you're just looking at the quality of the cars, then I'd say the Germans.

But maybe it's not quite that simple. I'm more concerned about who gets me what I want with the least amount of cost or effort.

However, a nice infrastructure doesn't necessarily mean you'll produce the best products. The Germans may have nice roads, but it's because the roads are heavily subsidized by taxes.

The French may have bad roads, but they cost less in taxes. If can just buy a car with good suspension you'll be ok. If you want to save money, you can deal with the bumpy roads.

And both countries have alternatives to cars: The excellent (but subsidized) rail systems.

Understanding

[Hnice]

Good article, but i'd nitpick over this:

"Of course, what's going on is not understanding, but logic, like you learn in high school"

Now, that's a stand you might take -- although i'd say that a meaningful majority of the people who think about these things for a living disagree. But the 'of course' is completely unwarranted -- this might be the most-discussed philosophical issue of the last 30 years, and it's dismissed here because apparently understanding means 'what humans do when they synthesize information, but not what machines do when they perform a very similar activity'.

like i say, this is nitpicking, maybe. it's a nice article. but i think that it's important, if we're going to make 'of course' statements about the relationship between syntax, semantics, and what understanding is, that we should remain cognizant of the fact that this is a terribly complicated issue without a whole lot of 'of course' about it. that is, i'm not clear on what grounds the author concludes that the semweb is not understanding.

Re:Heh

[Zeinfeld]

"Now XML has gone nowhere except as a set of popular libraries for cross-language data serialization..."

If there was a sentence in the article that proved it was rubbish this would be it. XML is not just slightly popular, it is now the defacto structured data representation. There is no competitor, there is simply no other format that is used in new protocol standards. Within ten years the DNS will have migrated to an XML format.

RDF on the other hand is a not very good idea to start with that has not exactly improved with the years. All RDF is in principle is typed set theory logic, so instead of trying to define a new set of semantics why not simply import Z or VDM wholesale?

Second problem with RDF is that it is really hard for a grad student to write an operational or denotational semantics for a programming language, a field that has only been worked on solidly for thrity years or so. So now we are expected to be defining semantics for everything???

The way that semantics get attached to syntax is through use. Use in this case means a program. I don't know that there is any RDF application out there that is likely to go much of anywhere soon.

I think that the way to get to a semantic web is completely different. You start from XML documents rather than attempt to change what the world chose for syntax. You build simple operational vocabularies of common terms for use in catalogues and make it really easy for people to categorize their work within those catalogues. You take as your starting premise that any structure of knowledge is going to be a work in progress.

Ouch!

[kavau]

The logic in the article is wrong. The example given,

"If A is a friend of B, then B is a friend of A,"

should read, as we all know, "If A is a friend of B, then B is a fan of A."

If they can't even get this simple logic right, I won't trust the rest of the article either.

Re:Heh

[mcrbids]

Within ten years the DNS will have migrated to an XML format.

I've heard some RETARDED statements on /. before, but this near takes the cake. DNS using XML?

Whatever you are smoking, I want some - 'cause it's clearly some REALLY GOOD SHIAT!

Given that:

1) DNS is a protocol, not a data format, and

2) XML is a data format, not a protocol, and

3) DNS is incredibly light and efficient, and

4) DNS has already proven that it scales well to just about any size, and

5) XML offers no particular advantage, since you could serve DVD ISOs over the DNS, and

6) moving to an "XML PROTOCOL" format would require the update of every single DNS server on the face of the earth, many of which are still running Bind 8.x, and some are still running BIND 4.X for god's sake,

I consider this to be HIGHLY UNLIKELY(tm) !!!!!