W3C home > Mailing lists > Public > semantic-web@w3.org > April 2010

Re: connections

From: Graham Klyne <GK-lists@ninebynine.org>
Date: Sat, 17 Apr 2010 16:27:51 +0100
Message-ID: <4BC9D377.2030603@ninebynine.org>
To: Lee Feigenbaum <lee@thefigtrees.net>
CC: Pat Hayes <phayes@ihmc.us>, Danny Ayers <danny.ayers@gmail.com>, Semantic Web <semantic-web@w3.org>
Lee Feigenbaum wrote:
> Seems to me that Danny's email is pretty straightforward: what is the 
> utility of all of these billions of triples of (primarily) linked data 
> that are being published on an almost daily basis? What can/are people 
> doing with them that are impacting lives or jobs or what have you? We've 
> made huge strides in publishing data, but it seems (to Danny, though I 
> can't say that I disagree) that useful ways to consume the published 
> data are still lacking. Which would lead me to question:
> 1) is that proposition true? if not, what are the useful ways to consume 
> linked data?
> 2) if the proposition is true, why? does it have to do with the quality 
> of the data? the content of the data? the access mechanisms available? 
> lack of problems to be solved? a dearth of UX or application design 
> perspectives that overlap with linked data awareness? is there a lack of 
> business models to drive people to build useful applications? ...

I think there's a social dimension to all this, and in my view people as a whole 
don't advance as fast as technology.

To the case in point.  Some very smart people have, as you say, made great 
strides to publish loads of data as RDF.  Now that it's there we need enough 
people to get used to the notion that it's there, to understand in broad terms 
the nature of what's available and hence the kinds of questions that
might reasonable be answered by consulting all this linked data.

For example, I think it took a few years for a significant number of people to 
get used to the idea that Google might be a useful early port of call when 
trying to find some arbitrary piece of information.

For linked data, it's harder, because we don't (yet, or ever?) have a single 
application that can be asked arbitrary questions to be answered by consulting 
the linked data cloud.  So when a new question arises, even if consulting linked 
data is considered, one also needs to judge whether it's worth creating a tool 
to find the data.  I see this something at a level similar to writing shell 
scripts:  do we issue a handful of commands by hand, or write a script (tool) to 
do it?  It certainly took me a while to get used to the idea of considering 
writing a new script (creating a new tool) for almost any new task that may come 

So, I think we want people to get used to the idea of consulting linked data, to 
start articulating requirements for tools for such, as well as some further work 
on tool building kits for rapidly prototyping such tools.  But getting enough 
people thinking in new ways will take some time.

Received on Monday, 19 April 2010 06:17:02 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:45:17 UTC