Embracing “Web 3.0”
Total Page:16
File Type:pdf, Size:1020Kb
Peering Editor: Charles Petrie • [email protected] Embracing “Web 3.0” Ora Lassila • Nokia Research Center James Hendler • Rensselaer Polytechnic Institute n an article published in The New York Times representation (KR), which is a subfield of artifi- this past November, reporter John Markoff stat- cial intelligence (AI) concerned with constructing I ed that “commercial interest in Web 3.0 — or and maintaining (potentially complex) models of the ‘Semantic Web,’ for the idea of adding mean- the world that enable reasoning about themselves ing — is only now emerging.”1 This characteriza- and their associated information. As such, we can tion caused great confusion with respect to the understand the Semantic Web through the lessons relationships between the Semantic Web and the learned from the Web’s development and adoption, Web itself, as well as between the Semantic Web as well as (perhaps somewhat painfully) from the and some aspects of the so-called Web 2.0. Some deployment of AI technologies. wanted to reject the term “Web 3.0” as too On the Web, we’ve seen the emergence of some business-oriented; others felt that the vision in the completely new business models that do indeed article was only part of the larger Semantic Web work, despite initially seeming infeasible. These vision, and still others felt that, whatever it was include the models introduced or perfected by called, the Semantic Web’s arrival in the Business Netscape (creating a community by giving stuff section of The New York Times reflected an impor- away), Amazon and eBay (marketplaces), and tant coming of age. Yahoo! and Google (advertising-supported sites). With the Resource Description Framework Sharing data (or content, as it’s often called when (RDF) and Web Ontology Language (OWL) — the discussing the Web) has unexpected and serendip- languages that power the Semantic Web — becom- itous outcomes — once you make something avail- ing standards and new technologies reaching able, you have no idea how some people will use maturity for embedding semantics in existing Web it. The long-tail phenomenon — for example, pages and querying RDF knowledge stores, some- aggregate sales of low-selling items, such as spe- thing exciting is clearly happening in this area. cialized books, surpassing the total number of best- sellers sold — defies traditional thinking about Semantic Web Background business models, but it’s important to the new Web- With more than 10 years’ work on the Semantic based economy. Web sites don’t really exist in iso- Web’s foundations and more than five years since lation — linking is what makes search engines work the phrase became popular, it’s an opportune and gives the “blogosphere” its power. moment to look at the field’s current state and From the euphoria surrounding AI in the 1980s future opportunities. From a humble beginning as through the hangover of the “AI winter” in the a methodology for machine-interpretable meta- 1990s, we’ve learned what doesn’t work: you can’t data and through a “world-embracing” vision of a sell a stand-alone “AI application.” These tech- new era of software (often — erroneously, in our nologies make sense only when embedded within opinion — attributed as science fiction), the other systems. Tools are hard to sell and often fail Semantic Web has matured into a set of standards to make good business sense (they certainly don’t that support “open” data and a view of informa- make sense according to venture capitalists). Final- tion processing that emphasizes information rather ly, thinking of AI itself, we observe that reasoning than processing. engines are a means to an end, rather than the end From one viewpoint, the Semantic Web is the itself; how you use them is more important than symbiosis of Web technologies and knowledge the mere fact that you use them. 90 Published by the IEEE Computer Society 1089-7801/07/$25.00 © 2007 IEEE IEEE INTERNET COMPUTING Embracing “Web 3.0” More recently, many people have siderable effort in revisiting and store data in a flexible schema so you can become excited about Web 2.0. Al- extending the functionality in the store additional types of information that though we abhor both the term and OWL standard, which is now emerging you might have been unaware of when you the use of version numbers, we see as OWL 1.1. originally designed the schema. The second that the movement is rife with inter- Given that much of the current is that it helps you to create Web-like rela- esting phenomena. Web 2.0 is mostly work was presented at academic con- tionships between data, which is not easily a social revolution in the use of Web ferences, that new journals have sprung done in a typical relational database. technologies, a paradigm shift from up relating to semantic technologies, the Web as a publishing medium to a and that much of the language design As RDF acceptance has grown, the medium of interaction and participa- happened in academic labs and corpo- need has become clear for a standard tion. From the Semantic Web view- rate research centers, some have under- query language to be for RDF what point, however, the most interesting standably assumed that the Semantic SQL is for relational data. The SPARQL technical aspects are Web is primarily a research vision that’s Protocol and RDF Query Language not yet ready for prime time. However, (SPARQL),3 now under standardization • Folksonomies (or “tagging”) pro- we’re starting to see considerable devel- at the W3C, is designed to be that lan- vide an organic, community- opment within the applications space guage. As Nova Spivack, CEO of Web driven means of creating structure and, as the “Web 3.0” article revealed, startup Radar Networks (www.radar and classification vocabularies; they often succeed where tradi- tional mechanisms for defining Web 2.0 is mostly a social revolution in the ontologies have failed or at least proven cumbersome. use of Web technologies, a paradigm shift • Microformats — the use of HTML markup to decode structured data from the Web as a publishing medium. (with the underlying thinking that human-readable representation now comes free) — are a step toward this work is emerging in an important networks.com), put it in a February “semantic data.” Although not in and exciting way. 2007 blog, “There is a huge amount of Semantic Web formats, microfor- interest in SPARQL at the moment, and matted data is easy to transform Web 3.0 there are already a growing number of into something like RDF or OWL for Although Semantic Web proponents SPARQL endpoints popping up around Semantic Web agents to process. have long seen evidence of growing the Web. These new SPARQL endpoints W3C is working on new approach- interest, the technology’s success has are to data what Web sites were to es, such as Gleaning Resource become far more evident in the past documents.” Descriptions from Dialects of Lan- few months. This is largely because of Numerous players of various sizes guages (GRDDL) and RDFa, to stan- the maturing of the RDF languages and are now focusing in different areas of dardize the linking of structured the technologies that support them. the Semantic Web space. UK-based data with instructions on how to Oracle’s July 2005 release of RDF Garlik (www.garlik.com), for example, transform or embed data into exist- support in its Spatial 10.2g database uses Semantic Web technologies for ing Web resources. product provided the legitimacy that the “control of personal data in the some felt the language lacked. As peo- digital world.” Specifically, the compa- Since the 2004 completion of the ple experimented with RDF databases, ny is working to let users discover RDF and OWL standards, we’ve seen a they found significant advantages over what’s known about them on the Web lot of experimentation (and confusion) traditional structured databases in to see what the aggregation of this regarding the right representation lan- many cases, especially with respect to information (exposed via an RDF guage to use for any particular appli- embedding data on the Web. As store) reveals. Dave Beckett, an engi- cation. Not surprisingly, subsets and Microsoft put it in its December 2006 neer at Yahoo announced in Novem- extensions of these languages have Connected Services Framework 3.0 ber 2006 that the Yahoo Food site started to appear — most notably, ver- Developer Guide:2 (http://food.yahoo.com) is being pow- sions of RDF(S) that borrow a small ered by OWL and RDF, as well as sev- number of features from OWL (though There are two main benefits offered by a eral other technologies. Teranode remaining simpler than “OWL Lite”). profile store that has been created by using (www.teranode.com), among others, is Other developers have invested con- RDF. The first is that RDF enables you to exploring the use of Semantic Web MAY • JUNE 2007 91 Peering Browser HTTP Ontology with SPARQL Back Forward Stop RefreshHome Print Mail @ AI application Address: > go Dynamic HTTP RDF often beyond your ownership or control content Dynamic RDF triple (and potentially hostile), requires funda- engine Code content RDF HTML and store mentally new approaches to ensuring engine triple reasoner interoperability. No longer can we Internet Zone store expect a priori standardization of every Figure 1. Sample “fractal” architecture of Semantic Web applications. Dynamic pair-wise interaction between all possi- content engines, backed by RDF triple stores, act as both producers and ble systems we anticipate interacting consumers of “semantic” data. Data exchange can be facilitated using, for with; indeed, we can’t even anticipate example, SPARQL.