Metadata data Metadata

Last week some of us were treated to a double dose of metadata, covering the topic in both the Digital Curation and 2D Digitisation lectures and labs. Despite it’s high-powered technical name, which might off-putting to the uninitiated, Metadata is really one of the simplest concepts in archival theory. It is data about data – the information about a record, it’s description, structural and administrative elements. What is not so simple is the complex mix of standards, schema and encoding that comes with it, beautifully expressed in a most colourful and complex diagram, A Visualization of the Metadata Universe.

In each lab session we had a go at creating metadata, from playing online metadata games, such as tagging British Library collections, to creating databases for hypothetical collections. However, what I found most useful and enjoyable was the real-life scenario of using and creating metadata for our Archives and Records Theory appraisal project. For that we got to experience the importance of Metadata first hand. Firstly, we had to find out as much as we could about the collection placed before us using the existing metadata, be it in a box list, accession register, catalogue or deposit file. Then we use our powers of detection to find out more information about our collection or associated collections in order to help us decide on the fate of the records. All the information we collect and the details of what is done with the collection and individual records is recorded as further metadata. This will ensure that future users will understand the context of the collection, it’s custodial history, what it contains and/or did contain (in the case of disposal or transfer to another archive) and ensuring transparency of the archival process clearly recording what was done and why.

Digital Curation Metadata Lab Questions

1. The amount of effort that would go into providing the full set of values for metadata elements in any collection. In particular, how much effort would go into encoding the full METS description (mets header, dmdSec, amdSec, fileSec, structMap, structLink, behaviourSec)
   1. I think most archivist would only put the minimum amount of metadata required to follow standards. Metadata can take a lot of time to complete, and it can be never ending.
2. The viability of automating metadata extraction. In particular, consider the variability across formats, also the crowd sourcing initiatives of the Metadata Games.
   1. I think its viable, although maybe not completely on its own. Someone would need to verify that there are no mistakes (such as tagging a cat as a dog), but I think it would help fill in metadata. As I said before, I think a lot of archivist do the bare minimum when it comes to data, due to a lack of resources, and automating metadata would help pick up the slack

Flashback to the 80s at the British Library

Today for our Digital Curation class former IMP student, Edith Halvarsson, gave us a really enlightening talk about the work she is doing as an intern with the British Library digital preservation team.  Edith has been working on the Flashback Project which is rescuing digital data from the 1970’s onwards.

By Trevor Johnson – Flickr: Retro Software, CC BY-SA 2.0

 Faced with the threat of data being lost forever as the storage media degrade and fail or the technologies become obsolescent, the project has been collecting floppy discs and CDs of software that is tricky to migrate, such as computer games.  Once images of the discs are captured, checked, and preserved on the British Library server, the testing can begin.  This is done in what they call the Legacy Lab, essentially a  room full of old working computers on which they can test out the original software.  They then use emulation software on modern computers to try to recreate the original gaming experience that was achieved with the old technology. The testing criteria look at the “significant properties” of content, context, structure, appearance and behaviour.  Any changes that have occurred through emulation are evaluated to see if they have a significant impact on the original message/aim of the software.  If the emulation has brought about unacceptable changes different emulation software can be tested out, but sometimes there is no solution and the game software has to be shelved until a suitable emulator becomes available.  The ultimate aim is not only to preserve these programmes but to make them accessible to users on computers in the reading room at the British Library.

The talk was really useful as it demonstrated the practical application of the theory of digital preservation that we have covered in lectures. It also tied in very well with the lecture we had earlier in the week from Matt Barr about his work on gaming studies at HATII. His talk looked at different generations of software and different emulations and the changes these brought to the user experience.  It is  quite scary to think how pressing time is for projects like this where old  media, like floppy discs, and the hardware to run them, are already failing.  Edith acknowledged that the Legacy Lab was only a short term solution and predicted that within 10 years they will have lost the expertise and replacement parts to fix the computers when they break.  As it is, they can only work on them for a maximum of 1-2 hours per day.

It was, however, very encouraging to see that a technological novice, like myself, could end up working in the digital curation field. Not only was Edith very engaging on the subject, she was also very enthusiastic about it.  Although, at the moment, my feelings about the topic are as far removed from hers as you can get, she gave me hope that one day I will see the light at the end of the technological black hole.