The Complexities of Metadata

In our MCPDM class this week we were looking at the values of standardised metadata. Metadata is essential for users to gain greater information about the object they are looking at, but we found out that compiling this information is incredibly complex.
Our first exercise was to examine the metadata in particular existing files. We could either do this by opening the properties of the file or uploading the file to http://www.extractmetadata.com. When right clicking for the properties of the file (depending on the file) we were able to view particular attributes, such as the date created, who created it and what format the file was in.
The website to upload your files gave some more information about the file that were not obvious in the properties, but I did not try all the files as I think we might have collectively crashed the website. Although both examinations of the files gave some aspects of metadata description, I still found it hard to find all 15 elements required for Dublin Core.I found that it was even more difficult to find the selected elements required for PREMIS metadata for each file.
However, looking at the metadata that was included for each file was very interesting, as it made me realise the diverse elements that can be included in file description. One image file even had the geographical coordinates to describe where the photograph was taken. Although having a few directional problems when trying to operate Google maps it was possible to find the exact location where the photographer was standing.
Our next task was to try to fill in metadata to particular files. We did this on the website http://www.metadatagames.org. This website had a series of games where you can tag and create metadata for various image and media files. The files were all provided by various organisations that could then use the tags and metadata the user had provided.
This way of collecting metadata is a good idea as it draws interest into the collections that are used in the games and it is a way of creating tags when institutions themselves may be limited in time and resources. I think it is particularly useful for creating general tags as many of the games had time limits and word limits. This is good for describing general aspects of the image or media themselves but not a good way of creating standardised metadata.
Some of the games allowed you to zoom in on the files to get a greater understanding of their content, but it was still difficult to find descriptions for the 15 Dublin Core elements. Some of the images I viewed were impossible to make appropriate metadata for. As some of the images were just showing a cloth book cover with no text written on it, which made it impossible to supply any information whatsoever.
Overall the exercises showed me how complex the formation of metadata really is. The varieties of elements that correspond with the variety of file types makes the various encoding of full descriptions time-consuming. Even finding the information to make up the elements required for particular standards can be extremely problematic. There is definitely more to metadata than meets the eye.

Advertisements