Tuesday, August 12, 2003

Why all this writing? #

Just a bit of explanation about the change in tone in this blog for those who want one:  Over the last month or so, I've discovered the great, great joy of writing in public, especially as a device for helping me to clarify my thoughts.  Hence, I'm jumping on the bandwagon in a big way.  I've decided to write more openly about my work in a as-it-happens fashion.  I certainly run the risk of further polluting the info-sphere with not-thoroughly-thought-out-thoughts and making a fool of myself.  However, the upside is you, my dear readers, get to see more deeply into what I'm learning and thinking about more quickly.  Clearly, one of the things I've discovered is that blogging and wikiing in no ways obviates the need for things like journals, which are places to hold more carefully conceived findings.  (Hence, my excitement for such projects as Pitch, which came back in my mind from today's post by David Wiley.)  However, there's plenty of stuff to share that is worth putting out there even if little of it is journal-level quality.   Posted by Raymond Yee on 8/12/03; 9:51:44 PM from the Personal Notes dept. Discuss

Jon Udell on Symbolic grounding and extensible aggregators #

Jon Udell's "Symbol grounding and extensible aggregators" will be useful reading when I get some time to think more about when/whether it's time to drink-the-RDF-cool-aid. It's probably close to that time, especially as we work on doing all these crosswalks between XML based formats (e.g., between METS and IMS-CP; MODS and IEEE LOM) -- and when we start to disaggregate and reaggregate content, mixing XML elements from different namespaces. I've been hoping that one will be able to have a sound strategy of mixing XML without having to go one more level of abstraction higher. Jon's article might shed light on this issue. (It will also be good to look again at Stephen Downes' take on Jon's piece.)   Posted by Raymond Yee on 8/12/03; 9:35:40 PM from the Web Technology dept. Discuss

METS vs IMS-CP -- an the limits of my current crosswalks #

Rick Beaubien and I are still slugging away at the METS/IMS-CP interoperability paper. Rick has done a great job on providing a detailed comparison of METS with IMS-CP in what we in a whimsical-wiki fashion called the BrainDumpRickBeaubien. The basic conclusion is that on the high level, there are basic similarities between the two specifications, but there are significant differences in the details of how the two specifications work. I'm now carefully reading through his analysis and comparing his analysis to the logic of the crosswalks I have written between METS and IMS-CP, for example, in "V.1" of the crosswalk (which transforms the structural aspects of the METS document but not the descriptive or adminstrative metadata). So far, I've found the following limitations for the crosswalk: ims:manifest/@identifier is currently hardcoded to MANIFEST01 -- is it appropriate instead to map mets:mets/@id to ims:manifest/@identifier? V1 does not account for the fact that mets:structMap is repeatable -- assumes only one mets:structMap V1 does not currently map mets:fileSec to imscp:organization V1 does not handle the following mets:structLink mets:behaviorSec mets:FContent The desire is to try to improve V1 of the crosswalk right now to address those limitations. However, we will resist doing so not only because we don't have the time to properly address these matters in the context of the paper -- but because doing so would take away from the essential simplicity of what we are trying to do. The reader is likely then to get immersed in the depths of not only one specification but two. It is helpful for me to remind myself what we are trying to get across in the paper in order to keep the focus of the paper strong. Although the general background is that of the library-educational technology interoperability problem (as laid out in the IMS-CNI whitepaper, for instance).the narrower scope of METS and IMS-CP interoperability is even beyond the range of the paper. Rather, we are focused on how library materials encoded in METS can be brought into a learning environments (learning tools) by translating them into the IMS-CP format (while preserving as much of the METS-related context) so that they can read in by IMS-compliant tools. There are no perfect translations for a number of reasons. First, the design goals of METS and IMS-CP are significantly different. Rick pointed out, for example, that it seems that the METS community is focused primarily on archiving and submission (SIP and AIP) but less so on dissemination (though I might be misunderstanding what Rick said) -- while IMS-CP seems clearly connected to the dissemination of learning materials. We both decided that it would be good to consult primary documents for both METS and IMS-CP to sort this out. Moreover, there will be context-sensitive (context-bound) learning metadata that will be determined in the context of creating learning materials from the "raw" library materials. No automated technique can determine such metadata. I'm writing these thoughts out so that I can be clear how I can dive back into the paper tomorrow....   Posted by Raymond Yee on 8/12/03; 9:23:33 PM from the Interactive University dept. Discuss

My infatuation with wikis #
Today, I wrote up on my personal blog some of my history with wikis and why I'm excited about them.  It's of relevance here because I've been using my wiki to do my professional work, including writing a paper.  (So much for the dichotomy between work and my personal life!)   Posted by Raymond Yee on 8/12/03; 6:09:33 PM from the Web Technology dept. Discuss