W3C home > Mailing lists > Public > public-lod@w3.org > March 2013

Re: Visualizing Linked Data - did we miss anything?

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Fri, 29 Mar 2013 10:43:04 -0400
Message-ID: <5155A878.4020401@openlinksw.com>
To: public-lod@w3.org
On 3/29/13 10:24 AM, Hugh Glaser wrote:
> Thanks.
> Yes, if your goal is "webby structured data", then what you describe is appropriate.
> Of course, others have other goals - this is particularly true of intranet applications using private data, etc.

Naturally. Thus, I won't even be able to access their intranets and 
wouldn't have anything to complain about :-)

>
> If/when organisations then decide that they want to share your goal (which we expect many will!), having used Linked Data as their internal platform technology will facilitate it.
> We should encourage the idea that Linked Data technologies can be used without being Open.

Sure, I've always interpreted "Open" to mean standards compliant. I've 
never interpreted it as meaning "Public Data". Thus, my concerns are 
only raised when documents are published on the World Wide Web (rather 
than networks based on Web technology) that offer presentation of data 
extracted from Linked Data sources that are already accessible via the 
World Wide Web.

> For example, Linked Data has a lot to offer in terms of multi-agency collaboration over sensitive issues (Child Protection?):- always banging on about Open tends to deprive such important fields of these marvellous technologies, if managers quite rightly worry about privacy.

Yes, and that's where WebID and WebID based ACLs that leverage relations 
semantics come into play. There are times when you need to provide 
controlled access (driven by data access policies) between different 
bodies, for sure.

> Sorry for getting a bit off the Visualisation subject.

It's on topic because there are also times when you want to drive the 
type and quality of visualizations based on the identity of an 
individual that's using a Web browser or some other user agent etc.. :-)


Kingsley
> Best
>
> (I still don't understand what meaning of "fidelity" you are using. Perhaps you mean the ability to verify fidelity, in some sense?)
>
> On 29 Mar 2013, at 14:04, Kingsley Idehen <kidehen@openlinksw.com>
>   wrote:
>
>> On 3/29/13 9:49 AM, Hugh Glaser wrote:
>>> Hi Kingsley,
>>> I think I understand what you are asking for, although I can't work out what "fidelity" might mean if it is being lost. (And I think the BM is very happy with where this little service fits in their value chain.)
>> Linked Data fidelity.
>>
>> The goal is webby structured data integrated into the innermost sanctum of the Web via URIs that have the dual effect of denoting things and identifying locations of the description of the descriptions of the things they denote [1][2].
>>
>> Unfortunately, in an attempt to hide raw URIs, many Linked Data visualization efforts end up throwing out the baby with the bath water i.e., URIs get discarded rather than being associated with labels using <a/>  or <link/> relations within <head/>.
>>
>> The pattern above enables discovery and exploitation of Linked Data URIs via user agents (e.g., browsers) and humans (who can then mouse over text anchored with a Linked Data URI based hyperlink or just click).
>>
>>> I can give you URIs, but they won't help you, as they are not to data that the organisation currently wants to make public by that method.
>>> (Eg http://bm.rkbexplorer.com/id/merlin-RRI34751 ).
>> The organization in question should keep the Linked Data URIs its consumed from the LOD cloud (and any other public Linked Data sources) discoverable to user agents and humans, as per my comments above. Of course, they can do whatever with their private Linked Data.
>>
>>> And of course you can't see the URIs in the web page, because it is a web page, not a Linked Data thingy.
>> No, the pattern I am suggesting is all about making Linked Data discoverable from basic Web pages where said pages are providing a presentation of data accessed from public Linked Data sources.
>>
>>> This is how it should be - the end user should not be exposed to the implementation technologies, and seeing a URI of a component of their user experience is certainly not going to help them.
>> That's not my point. As I've stated in the past, this is about the best of both worlds i.e., "horses for courses" compliance :-)
>>
>> Links:
>>
>> 1. http://twitpic.com/cdfva5 -- DBpedia URIs and explicit indirection
>> 2. http://twitpic.com/cbk8ul -- hash URIs as implicit indirection.
>>
>>
>> Kingsley
>>> Best
>>> Hugh
>>>
>>> On 29 Mar 2013, at 13:20, Kingsley Idehen <kidehen@openlinksw.com>
>>>   wrote:
>>>
>>>> On 3/29/13 9:13 AM, Hugh Glaser wrote:
>>>>> Hi.
>>>>> As I said, great initiative.
>>>>> Do you have a section or chapter about where Linked Data has delivered an enhanced user experience to existing web sites, rather than providing the whole experience?
>>>>> This is an important aspect for the eventual utility of Linked Data, although hard to capture.
>>>>> The sort of thing I mean is, for example, the "Research" and "Conservation" tabs at
>>>>> http://www.britishmuseum.org/research/search_the_collection_database/search_object_details.aspx?objectid=117631&partid=1&searchText=Rosetta+Stone&numpages=10&orig=%2fresearch%2fsearch_the_collection_database.aspx&currentPage=1
>>>>> come directly out of a Linked Data world (mediated by a sameAs store).
>>>>> There is quite a lot of similar stuff around, and I am guessing that educational resources would want to embrace that, and even discuss best practice.
>>>>> Best
>>>>> Hugh
>>>> Wondering aloud: are there any routes to the Linked Data URIs from the original Linked Data sources? It's a loss of fidelity when original source Linked Data URIs are disconnected from the value chain re., data visualization.
>>>>
>>>>
>>>> Kingsley
>>>>> On 29 Mar 2013, at 09:57, Maria Maleshkova <maria.maleshkova@kit.edu>
>>>>>   wrote:
>>>>>
>>>>>> Dear all,
>>>>>>
>>>>>> visualisation is obviously a very hot topic currently and there are a lot of tools and implementations, which provide different level of support. Some simply do a graph visualisation based on the links, other provide multiple visualisation forms to choose from.
>>>>>>
>>>>>> What I will try to do while preparing the chapter and the catalog is to identify the different visualisation needs that each of the tools address (simple browsing, exploring hierarchies, identifying relationships) . Furthermore, it is obvious that particular types of data are better visualised in a certain way (geo-spacial data --> maps).
>>>>>>
>>>>>> Naturally, the collection of tools should be available in an annotated way.
>>>>>>
>>>>>> As Barry mentioned, in creating the different chapters, we are trying to pick only the corresponding supporting technologies and tools.
>>>>>>
>>>>>> Maria
>>>>>>
>>>>>
>>>> -- 
>>>>
>>>> Regards,
>>>>
>>>> Kingsley Idehen	
>>>> Founder & CEO
>>>> OpenLink Software
>>>> Company Web: http://www.openlinksw.com
>>>> Personal Weblog: http://www.openlinksw.com/blog/~kidehen
>>>> Twitter/Identi.ca handle: @kidehen
>>>> Google+ Profile: https://plus.google.com/112399767740508618350/about
>>>> LinkedIn Profile: http://www.linkedin.com/in/kidehen
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>
>> -- 
>>
>> Regards,
>>
>> Kingsley Idehen	
>> Founder & CEO
>> OpenLink Software
>> Company Web: http://www.openlinksw.com
>> Personal Weblog: http://www.openlinksw.com/blog/~kidehen
>> Twitter/Identi.ca handle: @kidehen
>> Google+ Profile: https://plus.google.com/112399767740508618350/about
>> LinkedIn Profile: http://www.linkedin.com/in/kidehen
>>
>>
>>
>>
>>
>
>
>


-- 

Regards,

Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen







Received on Friday, 29 March 2013 14:43:27 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:16:31 UTC