Slashdot Mirror
Semantic Web Under Suspicion
Dr Occult writes "Much of the talk at the 2006 World Wide Web conference has been about the technologies behind the so-called semantic web. The idea is to make the web intelligent by storing data such that it can be analyzed better by our machines, instead of the user having to sort and analyze the data from search engines. From the article: 'Big business, whose motto has always been time is money, is looking forward to the day when multiple sources of financial information can be cross-referenced to show market patterns almost instantly.' However, concern is also growing about the misuses of this intelligent web as an affront to privacy and security."
"Semantic web" might make it easier for HackerBillyBob to find a potential identity fraud victim's information. So basically, HackerBillyBob can get dumber and dumber but do more and more damage. Fortunately the good side of this is PhisherBillyBob can decrease his R&D time because SemantiGoog will give him thousands of ACTIVE email addresses EACH AND EVERY MORNING.
Funnypics
So I know a lot of people that get all excited when they read articles on the "semantic web."
... you start with a corpus of documents to "seed" it and have experts (or web surfers) contribute little by little until it is accurate. Oh, it must also be able to adapt quickly and stay current.
I think that we are all missing some very important aspects of what it takes to make something capable of what they speak of. In all the projects I have worked on, to create something geared toward this sort of solution, you need two things: training data & a robust taxonomy.
First things first, how would we define or even agree on a taxonomy? By taxonomy, I mean something with breadth & depth that has been used and verified. By breadth I mean that it must be capable of normalization (pharmacetical concoctions, drugs & pills are all the same concept), stemming (go & went are the same action, dog & dogs are the same concept) and also important is how many tokens wide a concept can be. By depth I mean that we must be able to define specificity and use it to our advantage (a site about 747s is scored higher than a site about airline jets which is scored higher than a site about planes). By rigorous I mean that it must be tried and true
Without a taxonomy, how will we index sites and be able to tell between "water tanks" and "panzer tanks." I think that this is one of the great things that Google is missing to really improve its searching abilities. If you suggest an ontology to replace it, the problems encountered in developing it only multiply.
Where is the training data? Well, one may argue that the web content out there will suffice as training data but I think that more importantly, they need collections of traffic for these sites and user behavioral patterns to quickly and adequately deduce what the surfer is in need of.
I feel that these two aspects are missing and the taxonomy may be impossible to achieve.
Why are we even concerned with security if we can't even lay the foundations for the semantic web? I would argue that once we plan it out and determine it's viable, then we concern ourselves with the everyone's rights.
My work here is dung.
What I really want to see is the search engine reduce the duplicated content to single entries (try Googling for a Java classname and you'll see how many Google-searched websites have the API on them), or order them by reoccurrance of the word or phrase giving the context more value than the popularity of the page.
End the FUD
I can think of at least a dozen chain drug stores who have a retail store evey 10 blocks in every major city in the country. Their sales are recorded centrally.
Hypothetically, if all of them decided it would be for the good of humanity to allow someone to examine their sales in real time as a whole to identify flu outbreaks early - then the process of doing that would not be too difficult.
UPS and Fed Ex track their packages in real time, know who sent them and who is receiving them and how much it weighs.
Data about us is being collected more than I'd care to think about, and I think its inevitable that it become centralized. We may be rapidly approaching a point where we decide what we want, privacy or technology. Having both doesn't seem like its a viable option for very much longer.
Say what? Sounds similar to what the Bush administration via the NSA is doing only on a public level. For those who want to ramble on about privacy and abuse take note that just about every other week some company has lost their records, or someone has infiltrated their networks and gained access to records. If that's not enough to make you throw in the towel when it comes to protecting your privacy, never ever apply for a credit card, never sign a document, never reveal anything about yourself to anyone. What most fail to realize when complaining about these things is that hardly anyone takes the time to read the Terms of Service agreements. Whether its via purchasing something online, or paying your T-Mobile bill. If one did, one would take note at craftily worded crud a vendor created spelling out how they plan to share your information with others. So who is to blame when you bought something from Company A who sold your info to Company B and Company B loses your info or dishes it out. Or... Who will you blame when Company A who promised not to disclose your info is bought by Company B who made no promise to... As for privacy, get over it... Its diminishing slowly because you the people are allowing it.
Infiltrated dot Net
OK, I'm a bit confused by this statement from the article:
Depending on which sources the semantic web references, it could gather together health records, lists of recent purchases or even contact details, that would build a more complete picture of a person than ever before.
"All of this data is public data already," said Mr Glaser. "The problem comes when it is processed."
Huh, health records being public? Since when? Wouldn't that violate HIPPA? And AFAIK, retailers would never make purchase information free for the taking, so it's hard to imagine them allowing a search engine to start mining that data (not that they couldn't work out a deal were they get a piece of the ad revenue that a search may produce).
"I personally fear the day that a machine or algorithm can better determine the purpose for my keyword-based search than I can."
I wonder how this will influence our language and communication in general. Can language itself (not only its use) assimilated by marketing?
I shudder at the thought of 'marketspeak'...
...and growing and evolving.
Take a look at the "blogosphere" and the tagging/classification initiative that's happening there.
Sure, it seems crude and unrefined but it's working, like most grass-roots initiatives do when compared with grandiose "industry standards" and the big, bulky workgroups that try to define them.
body massage!
Obligitory Skynet joke in
5...4...3...2...1
-- 3 events that reshaped the world in the 20th century: WW1, WW2, and WWW
"Big business, whose motto has always been time is money"
That motto is really "anything for a buck". Even if business has to wait or waste time to get money, it will wait until the cows come home - then sell them.
--
make install -not war
The article would have us believe that this is going to expose everyone to massive amounts of privacy invasion. This is not necessarily the case. It is already the case that there are privacy mechanisms to protect information in the SW (e.g. require agents to authenticate to a site to retrieve restricted information). Beyond simple mechanisms, there is a lot of research being conducted on the idea of trust in the semantic web - e.g. how does my agent know to trust a slashdot article as absolute truth and a wikipedia article as outright fabrication (or vice versa).
As for making the content of the internet widely available, some researchers feel this will never happen. As another commenter noted that it is essential that there is agreement in the definition of concepts (ontologies) to enable the SW to work (if my agent believes the symbol "apple" refers to the concept Computer, and your agent believes it refers to "garbage", we may have some interesting but less than useful results). I am researching ontology generation using information extraction / NLP techniques, and it is certainly a difficult problem, and one that isn't likely to have a trivial problem (in some respects, this is goes back to the origins of AI in the 1950's, and we're still hacking at it today).
For some good references on the Semantic Web (beyond Wikipedia), check out some of these links
You could already do this semantic web nonsense if people would just stick to a standard and be honest with what they publish.
Nobody wants to do that however. Mobile phone companies always try to make their offering sound as attractive as possible by highlighting the good points and hiding the bad ones. Phone stores try to cut through this by making their own charts for comparing phone companies but in turn try to hide the fact that they get a bigger cut from some companies then others.
It wouldn't be at all hard to set up a standard that would make it very easy to tell what cell phone subscription is best for you. Getting the companies involved to participate is impossible however.
This is the real problem with searching the web right now. It wouldn't be at all hard to use google today if everyone was honest with their site content. For instance, removed the word "review" from a product page if no review is available.
Do you think this is going to happen anyday soon? No, then the semantic web will not be with us anyday soon either.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
How many people read this and thought "Okay, what have they done with Norton now?"
Find environmentally and socially responsible products on http://buy-right.net
The term "semantic web" generally refers to Tim Berners-Lee's notion of "semantic web" -- he coined the term. TBL's vision is simply a model of describing information. One expression of it is RDF-XML, another is data in N3 notation, but the core of it is the idea that you can express most information as a simple triplet of data: subject-predicate-value (e.g.: the sky - is - blue). That's basically it. It doesn't even have to have anything to do with "the Web" in the sense of the Internet.
The idea, however, is that you can represent information, and even information about that information, as a graph (where the object is itself the subject of another triple). If everything is uniformly presented, there's a slew of common operations you can perform on it, and merging data, making inferences from it, and such becomes easier.
The "semantic web" could be used to represent social networks (the apparent use case that is cause for hysteria), but it is not necessary and might not even be the best approach. Certainly, "semantic web" doesn't equate to social networks. For me, the greatest value is a method to store information about biological systems -- but then that's the field I work in. It could be used to add useful metadata to the web, it could be useful for making taxonomies and ontologies, all sorts of things.
However, the "semantic web" is itself about as sinister in nature as plain text.
Although I'm not particularly paranoid, I have no loyalty cards, only one credit card, frequently pay with cash and never borrow. In the UK, other people are registering cars at accommodation addresses so they cannot be located by speed trap and congestion charge operators, renting to avoid local government records, and generally finding ways to disappear.
Pining for the fjords
The guy claims that health records are public data? Well, that's a BBC site, but in the U.S. they decidedly are not, since HIPAA was passed.
But all this semantic web stuff makes me giggle when they start talking about healthcare, anyway. I worked in that industry up until a couple years ago. Semantic web people want to move everybody away from EDI...while the healthcare people are struggling to upgrade to EDI. In 2003 I was setting up imports of fixed-length mainframe records. By the time healthcare is exchanging RDF over the Web, we'll all have nanobots in our blood and won't need doctors anymore anyway.
The next great leap in searching the web won't be due to the semantic web. It'll be natural language processing. Soon the day will come when you will be able to type in a "real" question and truely get the best answers back. We all know keyword searching doesn't cut it. But a complete question can be interpolated to a logical query. It'll require no change to current web pages. Just a much smarter search engine.
Developers: We can use your help.
All the hoopla around the Semantic Web reminds me exactly of the days "XML" became the latest high-flying meme touted by "tech" writers en masse. Witness:
The semantic search engine would then cross-reference all of the information about hotels in Majorca, including checking whether the rooms are available, and then bring back the results which match your query.
And here in all its glory is the 1999 version:
The software would then use XML to cross-reference all of the information about hotels in Majorca, including checking whether the rooms are available, and then bring back the results which match your query.
Of course, the problem with this fantasy of XML was that no standardization of schemas led to an infinite mix of tagging and thus, the laypersons idea that "this XML document can be read and understood by any software" was pure bunk.
Granted, the semantic web addresses many of these problems, but IMHO the underlying problem remains: layers of context on top of content still need to be parsed and understood.
So the question remains: will the Semantic Web be implemented in a useful fashion before some develops a Contextual Web Mining system that understands web content to a degree that it fufills the promise of the Semantic Web without additional context?
Disclaimer: I work on contextual web content extraction software so yes I may be biased towards this solution, but I really think the Semantic Web has a insanely high hurdle (proper implementation in millions of web pages) before we can tell how successful it is.
Thanks for the information. A quick search for the W3C references on OWL fleshed out alot of what you were saying, and let me know know where I was getting off track.
The semantic web would have to be feasible before it posed some sort of threat, so I wouldn't get too up in arms about this.
concrete5: a cms made for marketing, but strong enough for geeks.
I've been trying to look for practical applications of Semantic Web. Can't find them anywhere. So far, what W3C proposes is a very high-level language. Of course, DAML,OWL... have nice features like cardinality constraints, axioms and so so forth. The idea is to build more logic into webpages. This is all theory. In practice, I don't really see how this will help making the web smarter. You can't expect people to write OWL by hand--like in the early days of HTML. We still lack an automated way of building taxonomies, and deriving the document's context and logical links with all other WWW documents. That's anyway the purpose of search engines, and so far Google is doing a pretty nice job at it. The next step is to build The Inference Engine, moving closer to A.I. Obviously we're not there yet. I think People who have the answer to this would reach the holy grail of computing. Not too mention being super rich!
At first glace I thought this was about Semitic Web.....
Now THAT would be something.
I judt got a nre Kinesis keybiartf so please excusr ant egregiou typos.
this is anti-semantic. does the ADL know about this ?
Sanity is the trademark of a weak mind. -- Mark Harrold
Oh good. They are finally going to include anti-virus on the web.
"All of this data is public data already," said Mr Glaser. "The problem comes when it is processed."
The privacy and security concerns are bizarre. They're saying that there is currently an implicit "security through obscurity" and that's ok. However, if someone were to make available data more easily found, then it would be less secure?
Here's a radical thought; don't make any data public you don't want someone to see. Blaming Google because you put your home address on your blog and "bad people" found you is absurd. If data is sensitive it shouldn't be there now.
You can't really bitch about peeping Tom's if you built the glass house.
I sent a terminator unit back in time to kill all the Slashdotters who were about to make an obligatory Skynet joke.
I am both a sceptic and a fan of the SW. I dislike XML serialization of RDF (and RDFS and OWL) - to me the SW is a knowledge engineering task and frankly 20 year old Lisp systems seem to offer a friendlier notation and a much better working environment. If you are a Lisp-er, check out the (now) open source Loom system that supplies a descriptive logic reasoner and other goodies.
The Protege project provides a good (and free) editor for working on ontologies - you might want to grab a copy and work through a tutorial.
I think that the SW will take off, but its success will be a grass roots type of effort: simple ontologies will be used in an adhoc sort of way and the popular ones might become defacto standards. I don't think that a top-down standards approach is going to work.
I added a chapter on the SW to my current Ruby book project, but it just has a few simple examples because I wanted to only use standard Ruby libraries -- no dependencies makes it easier to play with.
I had a SW business idea a few years ago, hacked some Lisp code, but never went anywhere with it (I tend to stop working on my own projects when I get large consulting jobs): define a simple ontology for representing news stories and writing an intelligent scraper that could create instances, given certain types of news stories. Anyway, I have always intended to get back to this idea someday.
The ideal Semantic Web is a beautiful dream, nothing more. Sure it'd be nice to have it around to start worrying about the consequences regarding security, privacy and everything else, but the ideal conceptualization seems to be impossible to implement.
The Semantic Web I see coming is one where many different, limited domains are identified and semantically annotated, allowing some kind of agents to perform well-defined activities for us (i.e., book tickets, make appointments, search info, etc.). This sounds more reasonable but although feasible, it is not an easy thing to do, for several very interesting challenges arise. Hence, the task of the research community should be to make this limited Semantic Web possible.
The "Semantic Web" is basically the intersection of RDF+OWL, that is to say, it is entirely about taxonomy. The whole idea is that you have a certain nomenclature that you assert against known values, someone else has a different nomenclature that they assert against the same values. You can now cross-reference with a high degree of confidence. For example, using the Dublin Core.
I get people all the time dismissing the whole idea because "man, you'd have to agree on definitions" or "how does 'it' know?" Right. "It" doesn't unless it is explicitly told. If what you call a "House" is in a well-known schema, you simply add an equivalency in your schema et voila, une maison est une 'House.' So, someone else comes along and they want to assert that "'ein Haus ist ein 'maison'," so they assert against the previous schema, and now implicitly ein Haus=une maison=a House. No one had to make the last assertion as it was implicitly true from the previous assertions. So now, in your schema, you make all sorts of categorical assertions about other things relating to houses. Your French and German counterparts now have them for free, as do you theirs. Yes, it takes work, no it isn't completely automatic, yes it is limited to strict taxonomies, but it is still very, very powerful.
The thing is, SKYNET was a military based computer and it gave us "Judgement Day".
..
I dare hypothesize that if a truly intlligent web ever arose, it would have a strong porn background.
I shudder to think of what it's version of Judgement Day would be
The sea changes color, but the sea does not change.
All this cross-referencing of data for the purpose of data mining (this appears to be simply a refining of data mining) is why I lie my butt off to telemarketers. Only by filling their collective coffers with garbage can we hope to undo their efforts. Currently, I'm a white male computer engineer when doing politically correct things, I'm a hispanic female for others, and I have many other identities. I pick identities at random for activities that I don't want traced, and I keep them active by doing strange things for the hell of it. Incidentally, I've been doing this since the mid-seventies when I registered for college as a black man. I changed races several times in college, as black student groups kept insisting that I join up. They get downright pushy about it. I think I ended up an American Indian - oops, Native American. What's it called this week?
I hope you people all understand that books, magazines, and newspapers are just as dangerous - anyone can publish private information using these technologies!
Come on, this is absurd. If anything this article underscores the need for privacy laws - but the privacy implications of the semantic web are hardly any more significant than any other publishing method.
The key problem is that we're bound by the inherent limitations of HTTP/HTML, which were developed as generic mechanisms for serving hypermedia. Within the parameters of hypertext, free text search with a nifty ranking mechanism is about as good as it gets, and Wikipedia is the killer app.
We need new protocols for richer information. How about a multimedia interaction protocol? Commerce/finance could probably use its own protocol. An "office productivity" protocol would be better than current efforts to shoehorn the MS Office filetypes into hypertext browsers. Machine-parseable semantic information may be a good extension of HTTP/HTML, but the waters have been muddied by the overuse of those protocols for information that doesn't really fit.
A heterogeneity of protocols and clients, customized for information types, would reduce the need for monolithic security fixes. It would be good for the software industry and -- in the long run -- good for users. The hypertext browser, with all its interface assumptions, has somehow been turned into the do-everything program and that model is not serving anyone well.
"Big business, whose motto has always been time is money, is looking forward to the day when multiple sources of financial information can be cross-referenced to show market patterns almost instantly"
That's the Bloomberg service in a nutshell. Yes, same company founded by current NYC mayor, Michael Bloomberg. As an example, I was able to simultaneously examine various financial ratios of about 1,200 companies along with their current market values. Depending upon where certain ratios went, I flagged them as candidates for a mailing of my company's financial consulting services. I was even able to use the Bloomberg system to download the names and addresses (office, not home of course) of the CEO and CFO of each company I flagged as a mailing list candidate.
Bill Clinton: Pimp we can believe in. - The Shirt!!!
The huge, glaring issue with the Semantic Web idea that I see is: how do you force the creators of web content to put the right semantic tags on their content? What's to stop there being thousands of sites full of nothing but semantic tags so that even Swoogling for "747" brings up porn first? The clear answer is that the tags will have to be out of the control of the creators of the web content. That means somebody or someTHING else - namely, your Semantic Web search engine of choice - has to figure out your site's tags for you. And the ONLY way to accurately judge, classify and rank a web page is by its actual, real content. This is just another way of looking at the same problem. I'm waiting to be impressed.
qntm.org
There's still people out there doing that stuff? That's too much! Good luck, semantic web dudes!
N.B. The above is a flippant, snide, and unhelpful comment. However, in my defence, I submit that that is _exactly_ the sort of comment that any remaining semantic web diehards should be most used to hearing.
Whence? Hence. Whither? Thither.