Shared publicly  - 
 
A semantic web value proposition challenge

In an earlier post today (http://bit.ly/LNpRXm) I quoted from a recent post by +Rob Gonzalez (http://bit.ly/O5dDuH) in which he said "Semantic Webbers have done a poor job communicating our value clearly and concisely."

This sparked an exchange between +Jason Pontin (editor in chief and publisher of Technology Review) and myself about the semantic web.

Jason:  No one - no one, including Sir Tim - can explain what the semantic Web is and does.

Me:  Nah, it's easy: the semantic web provides a way by which machines can talk to one another without pesky human intercession. ;)

Jason:  OK. I got that. It frees the Web from the document-pages-that-humans-click metaphor. But for what? What could the Web do then?

Me:  I guess I can't do better than the pat answer: make useful connections between things on the web.

Jason:  You see, this is where I get lost. The examples that I get never seem sufficiently compelling or new.

Which hearkens right back to the post by Rob that started this:

_"So question to the community: in one line, what is it that makes Semantic Web special?"

To rephrase this a bit (and finally get to the challenge to which I alluded to in the headline):  name a compelling example technology that demonstrates the usefulness of the semantic web.

I'll get the ball rolling by alluding to my brief Twitter response to Jason, which was "the Google Knowledge Graph".  The appended image is of the Knowledge Graph output for the query "Stephen King" (thanks to +Kingsley Idehen for his post on "simple examples of pretty page URLs from Google's knowledge graph" -http://bit.ly/MPlfuc - from which the King example is derived).  Putting together all those pieces on the fly is pretty rocket science-y,  at least to me, and I think the connections Google is now making by employing semantic web technologies is pretty compellingly useful to searchers.

But hell, I'm a search marketer, so of course I'm going to point to a search-related example.  I'd love to see an example that clearly shows the value of the semantic web without needing to reference that tried-and-true entity Google.
2
1
Ani Lopez's profile photoRob Gonzalez's profile photoMatthew Brown's profile photoAaron Bradley's profile photo
8 comments
 
+Kendall Clark When you answered "Cheaper enterprise data integration" what about the cost of deploying that new technology?  What about the cost of acquiring developers conversant with this new technology?  And isn't this just an incremental improvement such as one would normally make in an enterprise data environment?  What's so radical about it?  And what does it allow a data-rich organization to do

I'm playing devil's advocate here, of course.  But these are precisely the sort of responses I have received in response to that particular value proposition.
 
I like it +Ani Lopez, but can you put that into a value prop of 25 words or less? :)
 
I think the BBC is doing a nice job with their musician profile pages:

http://www.bbc.co.uk/music/artists/4a00ec9d-c635-463a-8cd4-eb61725f0c60

Those are useful semantic web results from Wikipedia (dbpedia) and MusicBrainz open data to populate the musician's profile as well as key links to their other homes on the web. 

Compelling and useful to users who end up on these pages? Yes. Compelling to publishers who aren't going to pay a mint to rewrite and populate this data across a giant sea of musicians? Their P&L statements say Yes.
 
+Matthew Brown Thanks, getting there I think.  And you're rightfully following up on Kendall's point on Twitter ("utility in which context?").

But we're still not there, value-prop wise, in the way one could conceivably be with another technology.

"And this PC is better than my IBM Selectric because...?"

"Because you never need to use correction fluid again.  Because you don't need to put something on paper until you're ready.  Because you'll never need to buy or install another ink cartridge.  Because you can change fonts without physically changing your wheel. Because you can save your work and use it on another PC.  For starters."

"Oh."
 
Interesting, I'll have to think a bit more about an external example or technology that demonstrates the value prop.

I feel like the crux of the issue is this. From your conversation:

"Me:  Nah, it's easy: the semantic web provides a way by which machines can talk to one another without pesky human intercession. ;)
Jason:  OK. I got that. It frees the Web from the document-pages-that-humans-click metaphor. But for what? What could the Web do then?"

Right now we've got

A. Enterprise Data Integration as in what +Kendall Clark means where the wins are largely internal and not necessarily public-facing 
B. Semantic web data smushed into HTML containers. Google Knowledge Graph. BBC Music and Olympics pages.

I'm not sure there's a typewriter -> PC example at the moment.
 
The issue I have with examples such as the Knowledge Graph, BBC Music, BBC World Cup, etc., is that they're not measurable results.  In any one of them it's hard to say that using Semantics was wildly more successful than not using semantics.  To me, the BBC World Cup example is very clear in terms of time to publish new content, but we never were shown analytics (that I remember) of increased engagement due to better links between subjects.
 
As far an increased user engagement due to semantic publishing on the World Cup pages, I'd love to see that stats on that. +Jem Rayfield doesn't look active on G+, but given the transparency they have with the DSP product, I'd think either he or John O'Donovan might be willing to provide some details.
Add a comment...