Comments on: There’s only metadata and URIs

By: Internet Alchemy » links for 2009-10-12

Internet Alchemy » links for 2009-10-12 — Wed, 22 Sep 2010 00:57:46 +0000

[...] There’s only metadata and URIs « Derivadow.com (tags: metadata linkeddata uris) [...]

By: Lego, Wombles and Linked Data « Derivadow.com

Lego, Wombles and Linked Data « Derivadow.com — Fri, 20 Nov 2009 09:24:10 +0000

[...] We often don’t give ourselves the best opportunity to womble with what we’ve got – to reuse what others make, to reuse what we make ourselves. Or to let others outside our organisations build with our stuff. If you want to take these opportunities then publish your data the webby way. [...]

By: Paul Rissen

Paul Rissen — Mon, 12 Oct 2009 10:25:32 +0000

Definitely. The story is a by-product (though that sounds a little harsh, to be fair) of the objects the journalist has identified, and the links he/she has drawn between them…

By: Tom Scott

Tom Scott — Mon, 12 Oct 2009 09:30:23 +0000

Not sure I agree – I think photos can be metadata.

If I have a URI for a person then with regard to that person a photo of them is metadata (information about them), although some people might feel a little odd describing a person as data. Or (since I work at the BBC) a programme page might have a photo to illustrate that programme (of piece of video), again from the perspective of the programme that photo is metadata about that programme.

Photos can also have their own metadata – but that doesn’t mean that photos should only be thought of as data. It’s metadata all the way down! Or data all the way down.

By: Paul Murray

Paul Murray — Mon, 12 Oct 2009 01:02:02 +0000

Seems that the data/metadata distinction happens when the “data” is in a different *language* to the metadata. An image, for instance, is not metadata. Perhaps the rule is: if google can pull it apart and understand the bits in it, then its metadata.

To put it another way, data is stuff that is *atomic* (as far as the web is concerned). Your point, then, is that today’s atom is tomorrow’s molecule. True enough, I suppose. An image that is a photo of a celebrity is not just an opaque set og bits, but with enough technology becomes a mention of that celebrity.

That being the case, “metadata” is simply stuff that fills in the gaps that the technology cannot do. One day, we will not need to accompany the photo with a caption saying “this is [President Barack Obama] at [the United Nations]” – the tech will extract it from the photo itself.

Until then, we need metadata for photos.

By: Tom Scott

Tom Scott — Sun, 11 Oct 2009 22:14:31 +0000

Agreed, it would also be nice if you knew something about the authors social graph, their beliefs, education, have they visited the country they are writing about (if relevant) etc. Find counter arguments and supporting evidence etc.

I guess the underlying point of my post was that until authors stop thinking of their story as the primary object and start thinking of the thing they are writing about as the primary object we’re never going to get there.

By: Tom Scott

Tom Scott — Sun, 11 Oct 2009 19:37:49 +0000

If some of the facts we currently ascribe to the person William Shakespeare (the concept) turn out not to be true then I don’t think most people would reject the William Shakespeare concept, nor would they wish to change that concept. However, our conceptualisation – the information we have about the concept William Shakespeare – has changed.

Likewise both you and I would probably agree there’s a concept/ a person known as William Shakespeare (or the place ‘London’ or the event ‘Wimbledon 2009′ etc.), however, we might disagree about how we conceptualise that person, place or event. That is a concept isn’t defined by it’s conceptualisation.

This is one reason why it’s important to separate identifiers for a thing, a concept from the information about it.

By: Paul Rissen

Paul Rissen — Sun, 11 Oct 2009 16:58:45 +0000

The news analogy is especially interesting to me – if you take a URI for an event, a URI for a story about an event, (and in that story, you link to all the URIs for concepts contained within the story, e.g. places, people, other events), *and* then a URI for the person who wrote the story, you potentially can start to build a web of people’s perspectives on events – and from that point, begin to draw interesting conclusions about prevailing opinions and interpretations of history and the world around us.

By: HB

HB — Sun, 11 Oct 2009 13:42:15 +0000

I’m not sure I understand what the last example is driving at, or your distinction between concept and conception. If WS is discovered not to have written the plays attributed to him (“English actor, lived in the 17th century, formerly believed to have written Romeo and Juliet and Hamlet, etc.”) – is the ‘concept’ the actual thing we’re talking about, in this case the dead dude called William whose history hasn’t been changed by this modern-day discovery, and the ‘conception’ our understanding of him, which is fundamentally changed?