The Structured Web and Linked Data are Leading to the Semantic Web
I have been a consistently vocal critic of “Web 3.0″ as a moniker for current Web trends, specifically as some sort of branding replacement for the semantic Web. My specific postings on the term Web 3.0, likely with lame attempts at derision, are on record as:
Then, in relation to an article on ReadWriteWeb regarding a keynote on semantics and advertising at last week’s Web 3.0 Conference & Expo in Santa Clara, CA, I added a comment joining others criticizing the use of version numbers to describe the semantic Web. My comment was simply: “Please, Squash that Web 3.0 Cockroach (see http://www.mkbergman.com/?p=406).”
I count Dan as a colleague and a friend and do not want to engage in a flame war across multiple sites. So, herein, I address his rhetorical question, “How shall we call it [Web 3.0] instead, Mike? Please indulge us.”
It just so happens I have been writing on this very topic for quite some time.
(BTW, while others commented on the proper role or not of semantics and advertising, that is not my issue. I have never criticized advertising or marketing and will not do so with respect to semantic Web techniques, either. I fully expect semantic technologies to be applied everywhere appropriate and where they can work, which most certainly includes better targeting of ads and messages.)
Getting Serious: Meaning is Important
So, what is my problem with the use of Web 3.0 for semantic Web-related trends and activities?
Simply answered, it is because Web 3.0 means nothing.
It can mean anything or everything depending on who is pushing the idea. I dare anyone who is pushing this term to find a consensus understanding or authoritative or “official” understanding or grounding in language or usage or any other informing basis for what this term means.
I find it ironic that the semantic Web, which is at heart about meaning and data interoperability, could potentially accept a naming or branding that is itself meaningless. Simply answered, that is my issue and has been since the term Web 3.0 was first floated.
Moreover, rejection of the term Web 3.0 does not carry with it the logical fallacy of therefore not wanting simpler ways to explain semantic Web-related concepts and technologies to the broader public. Nor does rejection of the term Web 3.0 carry with it the logical fallacy of not wanting effective branding, effective terminology or effective business models.
What rejection of the term Web 3.0 does mean is that we can and should do better in conveying what our collective endeavor is really all about. And with meaning.
So, Please Indulge Us
I have two answers to Dan’s rhetorical question: structured Web and linked data. I further place both of these concepts into context.
Many have given their own bona fide attempts at describing semantic Web aliases and where they stand within a development continuum. Some of this is the natural attempt by some to want to “name” a space and time. Some of it is undoubtedly a reaction to the glazed look many in the public get when “semantic Web” is first put before them. Some of it is likely a reaction to the fact that the semantic Web has been slow to develop or has not yet met its initial promise and (perhaps) hype.
Over the past two years, I have pointed to two concepts as part of an ongoing continuum eventually leading to the semantic Web. These are the broader and bridging structured Web, which is currently seeing one expression as linked data. I first tried to capture this continuum in a diagram from July 2007:
|Document Web||Structured Web||Semantic Web|
To further a common language, I have also put forward my own working definitions of these concepts:
- Structured Web is object-level data within Internet documents and databases that can be extracted, converted from available forms, represented in standard ways, shared, re-purposed, combined, viewed, analyzed and qualified without respect to originating form or provenance
- Linked Data is a set of best practices for publishing and deploying instance and class data using the RDF data model, naming the data objects using uniform resource identifiers (URIs), and exposing the data for access via the HTTP protocol, while emphasizing data interconnections, interrelationships and context useful to both humans and machine agents.
I have repeatedly discussed these themes and ideas since I first criticised attempts at the Web 3.0 branding:
- October 12, 2008 — WOA: A New Enterprise Partner for Linked Data
- July 25, 2008 — When is Content Coherent?
- June 23, 2008 — [*] What is Linked Data?
- May 23, 2008 — Structured Web v. Semantic Web?
- May 6, 2008 — The Semantics of Context
- March 12, 2008 — The Shaky Semantics of the Semantic Web
- February 3, 2007 — Linked Data Comes of Age
- July 22, 2007 — [*] More Structure, More Terminology and (hopefully) More Clarity
- July 18, 2007 — [*] What is the Structured Web?
- May 3, 2007 — Structure Paves the Way to the Semantic Web
- April 23, 2007 — Structurizing the Web with RDF
- April 5, 2007 — What’s in a Name?
- April 2, 2007 — Did You Blink? The Structured Web Just Arrived
- December 12, 2006 — The Pragmatic Semantic Web
Those entries marked with an asterisk [*] are the most central ones dealing with either structured Web or linked data as terminology.
Over two years, about one in six of my blog posts has been devoted strictly to meaning and terminology. I agree proper branding for our collective endeavor is very important. In the end, it is not important whether my views hold sway, but that we are able to effectively explain and sell our products and services. There is still considerable outreach and communication required with the marketplace.
The Markets will Decide
Amongst the two concepts, I prefer the terminology of structured Web because it seems to be readily understood and appreciated within enterprise clients. But, we have also embraced linked data because it was gaining mindshare and conveys (imo) a concrete and correct image.
Terminology adoption is a function of both providers and consumers. Consumers vote with their attention and their wallets; providers through their branding and positioning, which naturally should be attentive to what is resonant in the marketplace.
Right now, in the emerging markets around the semantic Web and its related technologies expressing the current evolution of this continuum, the naming issues are still largely in the purview of the providers. Consumers are still few. I think most would agree that terminology is still unsettled.
It is legitimate to question whether the provider community is doing itself a disservice using ‘semantic Web’ as a hook. It is healthy to seek better terminology if it can be found. Any attempt to find clear and compelling language is to be applauded.
But, insofar as we are selling improved meaning and data interoperability, let’s find language that conveys those advantages. Web 3.0 directly contradicts this fundamental message by offering no meaning in its vacuity.
Don’t Crush that Dwarf; Hand Me the Pliers
Well, come on now, let’s do smile a bit. There are worse things than calling the term Web 3.0 a cockroach. After all, cockroaches have existed for 340 million years well in advance of humans and can be found everywhere. More than 5,000 species have been identified. Some claim cockroaches will be here long after humans are gone.
Still, as for me, I’m hoping the term Web 3.0 is squashed well in advance of that time. Meanwhile, I appreciate the invitation to indulge in meaningful branding and terminology.