w3c_home_nb.pngWeb 2.0 is so yesterday. Perhaps, that’s an overstatement, but Web 3.0, the semantic web, is rapidly moving beyond the academic laboratory to real-world utilization. Now the W3C RDB2RDF Working Group has published the First Public Working Draft of "A Direct Mapping of Relational Data to RDF.” This work is a major step towards making the astounding amount of data stored in relational databases more easily available to the semantic web.

Semantic What?

Web 2.0 focused on establishing interactive human-to-human communities on the internet with rich content. Web 3.0 evolves beyond this and strives to make content consumable for people and machines. Semantic technology uses RDF, Resource Description Framework, a XML-based language to mark up content as structured data that computers can interpret.

For example, an article about me might indicate that Josette Rigsby writes for CMSWire. RDF allows you to indicate that the two topics, Josette Rigsby and CMSWire, are connected by the "employed" relationship or:

SUBJECT: Josette Rigsby
PREDICATE: employed at
OBJECT: http://www.cmswire.com

In RDF, the predicate is also represented using URI, but unlike objects, predicates are selected from an established list of URIs categorized by topic. For example:

Why Should You Care?

Ultimately, the semantic web will make it possible to query a web of linked data exposed by websites as if they were a database, the world’s biggest database, and follow the links between the data. This capability would make it possible to create rich experiences that are not easily possible today.

Much of the data on websites is stored in relational databases. In fact, according to the Association for Computing Machinery, over 70% of existing websites derive data from relational databases. Given this volume of information stored relationally, the promise of the semantic web requires making this content easily accessible.

This is precisely what the recent W3C RDB2RDF Working Group draft does. In progress since late 2007, the draft proposes a standardized language for mapping relational data and schemas to RDF. The mapping does not depend on the schema of the database -- the language defines a general mapping for any relational structure into RDF. Users specify base URI for the database and the rest is generated. Once this occurs, users can leave the data in the relational database and link it to web using the mapping.

Who’s On the Semantic Bandwagon?

The excitement and benefit of semantic technology isn’t just being written about in techie blogs. Gartner (newssite) has predicted that by 2017 the majority of websites will include some form of semantic markup. In July of this year, Google acquired Metaweb, makers of Freebase, Metaweb’s free and semantic database of over 12 million things. Last month, Extractiv, a service offering on-demand semantic conversion of unstructured data, announced their public launch.

The big boys of software are also taking notice. Oracle offers Oracle Spatial 11g, a separately-licensed RDF management component for its data platform. Cisco has not yet released a semantic product but is offering customer education. Microsoft is active in semantic search via their acquisition of Powerset and integration within Bing.

IKS has also released its first demo of a semantic knowledge engine and there are a number of content management vendors looking at incorporating this technology into their solutions.

Business verticals also see the value of the semantic web. BioPharma is perhaps the best example. The industry is adopting semantic technology to address data integration challenges -- companies like Eli Lilly have been using the tool for researching drug targets of interest.

As the web continues to evolve, providing improved access, and context for information grows increasingly important. Are you actively using semantic? Do you think this proposal will spur advancements? Let us know