7 Aug 2003 @ 21:07, by Roger Eaton
RDF and the Semantic Web are an intriguing development, built right on the same street where we are clearing a plot for InterMix. Investigating the Semantic Web led me to Chandler which looks even more promising.
RDF - Resource Description Framework and the Semantic Web. The Semantic Web is a project of the W3C, World Wide Web Consortium and has the imprimature of Tim Berners-Lee. See his seminal paper Semantic Web Road map. The RDF specification has been accepted as an official recommendation of the W3C.
The fundamental concept that underlies RDF and hence the Semantic Web is that of a resource/property/value triple. A resource is anything that can be indicated in words, such as a dog or an economics concept or a class of machines. A property is an attribute of a resource. (To see how things get complicated quickly, a property can also be treated as a resource -- after all, it can be indicated in words, and it may have properties of its own.) A value is the particular of this property applied to this resource. Resources in general have many properties. Thus, for example, you may have two triples, 1) country/trade-imbalance/half-a-trillion-USD and 2) country/name/USA.
RDF is a natural for laying out a table of a database. Each row in a table represents a resource, each column a property and each cell contains the value of that property for that resource.
The Semantic Web proposes to use RDF as a basis for making data on the web machinable. If everyone as a matter of course puts their web data in RDF/XML format as well as in standard HTML or XML, then anyone who wants to repurpose that data will have a much easier time of it. Much more high quality data will become available as organizations automatically pump out their databases onto the web in RDF and a new generation of search engines will arise to comb this shared world database for easy access by everyone.
A truly huge amount of effort has gone into the Semantic Web with very little return so far. The resulting accumulation of concepts is daunting and there has arisen a fear that the Semantic Web will never hit critical mass because, unlike html or even xml, RDF/XML is just too hard for your average astrophysicist. Some important guys are raising a big question mark. Tim Bray in May of 2003, The RDF.net Challenge, has urged the Semantic Web community to rethink (which doesn't seem like it is going to happen, because so much effort would have to be undone) and Marc Andreessen, who in April 2003 said "he doesn't think so". The fear is not so much that web designers will steer clear, but that web programmers, realizing that it will take days or maybe weeks to get their brains around this stuff, and seeing no visible result for their efforts, also won't bother.
Francis Heylighen has a thoughtful remark in an interview with nanomagazine: "I believe we are confronted with a complexity bottleneck, which will significantly dampen the speed of further progress. The human mind simply is no longer able to cope with the information overload. This also means that all the big software projects that require a lot of coordination between different people and sources of information (e.g. the present 'Semantic Web' efforts) either will get seriously delayed or end up with buggy products."
On the other hand, here is a recent article, "The Future of Human Knowledge: The Semantic Web" that is cautiously optimistic about RDF.
Still, for a major effort now five years on or so, the Semantic Web is not percolating the way one would expect if it is going to take off. Too much work for no immediate payoff, and no eventual payoff until/unless everyone adopts it. Chicken and egg.
The Semantic Web also smacks of "brittle" artificial intelligence and the still not successful CYC experiment that is attempting to code common sense rule by rule. But this is not really a fair comparison, since SW is not meant to be an intelligence. Rather it is meant to improve the access of data across the web, and it is, in theory anyway, open to development of the guiding ontologies, not locked into a particular schema. It is a hard problem, the categorization of knowledge.
In respect to the InterMix collective communication project, the Semantic Web may well supply ready made rdf resource/property vocabularies. See the Friend of a Friend project with its RDF vocabulary for a "Person". Or the w3c rdf adoption of ical. These are examples. As these vocabularies accumulate, they should be very useful as InterMix "dimensions". More about which later. And the other way about, the InterMix Voice for Humanity project will want to publish the "dimensions" that we develop as rdf vocabularies.
As a practical matter, InterMix will want to be able to import and export RDF. Very possibly open source RDF tools will prove useful in the development of the InterMix query language.
|
|