Monday, February 17, 2014

Tagging is a Complex Mistress


I've given a little bit of thought to the tagging schema of Arrian's The Campaign of Alexander. My first instinct was to begin tagging from left to right - I need to start somewhere, after all. And the archives and record management stream has beaten respect for original order into my brain so much over the last few months that I've started applying it to places in my life outside an archival setting. With this specific image, that would mean starting with the page title printed farthest on the left: "Battle is joined on Alexander's right". That would be followed by the section title: "Gaugamela", then "Autumn 331", then "Book Three". The second page in the image, from left to right, would be: "Book Three", then "Autumn 331", then "Gaugamela", then "Darius sends out his chariots".  The inconsistencies are self evident. The hierarchy of titles is unclear.   

The method of tagging titles as they fall on the page may work if these two pages were not considered part of a larger volume. But it's clear that method will not suffice for a number of reasons.  As you can see, there are four titles (and I use that term loosely) on each page. Only the titles in the green circles describe different content. It's impractical to repeat the yellow, red, and blue tags for every page. It also doesn't give any accurate context as to which title is the largest section; is "Autumn 331[AD]" larger than "Book Three"? In this case, not likely, but who is to say for sure without access to the physical book? The tagging schema will have to be broken down from it's largest parts to its smallest - an imposed hierarchy. Using the colour scheme below, that would be yellow, red, blue, then green. "Book Three" changes the less frequently than the other sections. The green circle's title changes, presumably, with each page. Consequently, the page layout will be lost in XML-translation, but the content will remain and users will be able to figure out what section of Arrian's text they are reading.       





There is another issue at hand with tagging this image. This issue may not be quite as easily solvable as the title problem.  The issue at hand is the descriptive notes in the margins.  In the red circles,  "Autumn 331" is used as one of the sub-titles at the top of the page and it is used again as sub-title and contextual element for the note in the margin.  In that note on the right side of the image, there is an overlapping blue and red circle that merges "Gaugamela" and "Autumn 331" as part of the same description.  Can these two descriptive elements become one tag? Further scrutiny of the text will help decide if tagging these names again as a contextual element would be redundant.  One option is to omit tagging these sections and leave it as regular content. Or in the interest of creating a user-friendly search aid, a distinction will need to be made between the sub-title "Autumn 331" and the contextual element "Autumn 331".  

Food for thought. 

No comments:

Post a Comment