Please enable JavaScript to view the comments powered by Disqus. comments powered by Disqus

Report #2 from a stray climber on the MDM Summit Europe 2016

Metadata Nightmares

I woke early this morning. On Sunday I’ll be travelling to London for the MDM Summit Europe Conference. My mind began churning metadata and I drifted into a dream.

I was at the Workshop about DAMA CDMP Certification by Chris Bradley on Monday morning. He showed his slide with a photo of his cats as an example of data and metadata. In my dream I flipped and began fulminating.

metadata of the image of a couple of cats

In Chris’ presentation this is an example of data and metadata. The image is data and on the right is, according to him, metadata about the image. This is data that is actually embedded in the image file. On Flickr you can click to reveal it for each photo. There is a long list of elements.

So in my dream my tirade started by pointing out that on the right I see two columns and one appears to be metadata and the other is data. My pink arrows. The data includes elements about aspects of the image file, the image itself, the camera and so on. The metadata is comprised of labels for for these elements, probably according to the EXIF standard. Although each camera manufacturer seems to have its own implementation of the standard. So to say that the image is data and all the stuff on the right is metadata is a bit of an oversimplification.

So I continued raving - really DAMA should get their act together, follow their own guidelines and develop a logical data model of this photo of cats. There would be several objects and their attributes in such a model:

  • image file
  • image
  • camera, distinguishing the body and the lens, if it is a serious camera
  • various people
  • cats
  • copyright or other licenses
  • place where the photo was taken
  • event or occasion - maybe it was the cat’s birthday
  • date and time and time zone the photo was taken
  • software used to edit the image

The EXIF data includes some text elements, title, description and keywords or tags. There are endless possibilities for filling this text with stuff about the photo. Really we need a model of the universe. After all such an image can contain anything real or imaginary in the universe. Cats aren’t the only thing in the universe.

The workshop leader protested. I was getting out of hand and off topic and disturbing the other participants. I had anticipated this, and responded with expressions of adoration for the DAMA DMBOK. I had been so excited when I discovered it. It includes some beautiful writing. The introduction to the concept of architecture is lyrical. But the chapter on metadata goes off the rails. It starts with the normal definition of metadata as data about data but it ends with a list of twenty or more classes of data that are said to be types of metadata. It seems even the cloud servers that the data floats on and the stewards who lovingly tend the data are all declared to be metadata.

And there is the woolly example of the library catalogue as metadata. In my dream I drifted off into Umberto Eco’s novel The Name of the Rose, set in an Italian monastery in the Middle Ages. At the heart of the monastery is the library, a collection of books built up over centuries, from many places and in several languages. The monks work every day at their desks in a big writing room, studying, transcribing and writing commentaries on the books in the library. But they are not actually allowed into the library. Only the librarian has access to the collection. For the monks the key to the library is the catalogue, a list of the books in the library in chronological order of acquisition.
Concept Map of objects and attributes represented in Exif Data

I’ve been reading it on the train, on my iPad, using the iBooks app. Every Wednesday I go on the train to Amsterdam to look after my grandson, Mozes. A train ride of an hour and twenty minutes. My dream became a nightmare. DAMA would want all that as metadata. The iPad, the train ride, the death of Umberto Eco as reason for reading the book. The association between Mozes and Umberto Eco. Everything becomes metadata!

Thankfully I woke up before I went mad and my mental condition had to be added to the DAMA list of types of metadata.

Often dreams slip away before you are fully awake. In this case I have tried to capture the objects and attributes represented in EXIF data in a rough concept map.

Looking forward to some calming words from Chris Bradley at the workshop on Monday morning…
blog comments powered by Disqus