W3C home > Mailing lists > Public > public-sweo-ig@w3.org > February 2007

Re: Open data linking community project.

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Thu, 15 Feb 2007 10:25:13 -0500
Message-ID: <45D47B59.7000708@openlinksw.com>
CC: public-sweo-ig@w3.org

Kjetil Kjernsmo wrote:
> On Wednesday 14 February 2007 19:29, Kingsley Idehen wrote:
>   
>> The Linking Data Project is in no shape of form Geek oriented. This
>> is unfortunately an incorrect characterization of the project.
>>
>> The last thing Chris, Richard, Stefano, myself, and others have in
>> mind is yet another Semantic Web Geek Project. We are all very
>> anti-semantic-web-geek. This is why we have specifically set a goal
>> to make this project usable by the hypothetical Grandma.
>>
>> Please send some specific questions about this project that can aid
>> in correcting the current misconceptions.
>>     
>
>   
Kjetil,

> Hi Kingsley!
>
> I was hoping you would chime in on the meeting, but I guess you had some 
> connection problems.
>   
I had more serious problems unfortunately on the family side :-( Things 
have settled now.
> On one hand, I think you do appreciate the problem of getting semweb out 
> to the masses at least as much as I do. Moreover, with all the 
> commitments you have, I believe the project stands well on its own 
> feet, and also merits a lot of attention from SWEO.
>   
Sure!

> But OTOH, I don't quite see how this project, in its immediate form, 
> will contribute to something that the average user can use. I feel that 
> Tabulator is not ready for the average user, but Tabulator is useful 
> for demonstrating key strengths of the Semantic Web.
Tabulator is not the definitive front-end of this project. It is one of 
several. The idea is to put a collection of applications (data 
consumers) and data sources (data providers) together within coherent 
whole that demonstrates:

1. What the Web's Protocol gives already via HTTP Content Negotiation
2.  Linking to other Data Sources via the common use link oriented 
predicates (such as foaf:seeAlso and foaf:knows plus others on a schema 
by schema basis)
3.  URI Dereferencing

The above are the heart and sole of the main problem re. the semantic 
web foundation layer (the Data Web part). They dispel the "RDF Tax" myth 
by unobtrusively exposing RDF in a myriad of ways.

>  Applications that 
> can allow average users to connect open data is clearly very good, and 
> very much in line with the idea "we've got a billion people out there 
> who can state that foo and bar is related" (as opposed to the 
> stereotypical ontology approach "foo and bar is related through the 
> ontology", the topic map "foo and bar is related because there exists 
> resources that has both foo and bar" etc ;-) ). But there is no 
> incentive for average people to participate in this unless there is an 
> application that does make extensive use of it.
>   

Which is what this project is aimed at doing hence the broad 
participation you see. There is no silver bullet for solving this 
matter, all we can do is provide a boat load of solutions that simply 
deliver choice to as broad an audience as possible without being 
distracted. This is a Many-to-Many assault on the issues relating to the 
RDF image paradox (those who dislike  RDF are the ones that benefit the 
most, they ask for RDF on a daily basis without knowing that RDF is what 
they are asking for etc..).

> And I don't quite agree that there are only toy examples out there, 
> there is like 17 million FOAF profiles, for example...
>   

I didn't say that :-) I assumed you gleaned this from some of the 
material one page? Naturally, the Many-to-Many nature of these projects 
inherently applies to the rhetoric :-)
> So, what I feel the answers to the questionnaire lacks is the 
> application that combines Wikipedia, MusicBrainz and Geonames into 
> something that will change my life. Something like "with these links 
> between open data, we will create a plugin for Amarok, XMMS and WinAmp 
> that allows users to find information about similar bands in the same 
> geographical region". And then, argue why this approach is superior to 
> just googling the names. 
>   
The one thing you are missing is that SearchTheSemanticWeb (lead by Fred 
Giasson) is about Grandma's Semantic Web experience (and there will more 
than likely be others the find their way to this effort).


> And of course, I'm really happy about your offer to host the data 
> source. I was a bit worried about where to host the data, as it will 
> likely receive quite a lot of traffic. And evidently, it will be 
> interesting to have linked data to the FOAF data. 
>
> But I don't quite feel that the answers to how it would become 
> widespread was adequate, and apparently other SWEO members agree. 
>   
They all of this becomes wide spread is via massive adoption via 
"network effects" one friend tells another friend about their Data Web 
experience which eventually leads to a global epiphany :-)

> Now, I wouldn't want to influence the SWEO members too much. It could 
> easily be that people think that "Kjetil created the questionnaire to 
> suit his own project, duped SWEO into condoning it and then went out 
> victorious". Which would be bad... 
>
> So, I'll shut up now but I hope you can elaborate on that.
>   

BTW - What does SWEO support of a project actually entail? I ask because 
(as I stated earlier) the Education and Outreach aspects of these 
projects is already exceeding expectations (from my vantage point). The 
collection of projects is gradually unveiling a less intimidating 
conduit to the semantic web realm. The same applies to practical usage 
scenarios, patterns, and best practices. These projects a basically have 
their own evolutionary paths and I think simply have SWEO exposed as the 
catalyst for this effort is good enough from the SWEO perspective.

I don't want us to inadvertently tapper the process that is in motion by 
associating SWEO with a specific project based on a voting process of 
sorts. The projects, or developers associated with a given project will 
easily misread the intentions here.

Personally, as I have stated in the past, the more projects the merrier. 
We just want to have lots of coherent sources of material that shed 
light on the practical benefits of the Semantic Web vision.


Kingsley

> Cheers,
>
> Kjetil
>   


-- 


Regards,

Kingsley Idehen	      Weblog: http://www.openlinksw.com/blog/~kidehen
President & CEO 
OpenLink Software     Web: http://www.openlinksw.com
Received on Thursday, 15 February 2007 15:25:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:35 GMT