Talk:Module 4-b: Metadata

From DL Curriculum Project

Body of Knowledge: Does the module address all areas of the topic that need to be addressed?

1. Will the body of knowledge enable students to achieve the objectives?

• yes --Hoangs 11 June 2008 (EDT)

• I am presuming the the word "remedial" implies that there has been a general concepts and XML exposure already. It's a little hard to tell from the general outline framework. The present body should enable students to attain the learning outcomes. --Schwartzc 09:56, 12 June 2008 (EDT)

• I am pleased with the range of topics in the body of knowledge for the module. I cover most all of these topics in my metadata course, and even have sessions on education metadata and the Semantic Web (though my coverage of these topics are just enough to pique the interest of students in taking a yet to be designed, more advanced elective course). -- Maccalls 14:01, 12 June 2008 (EDT)

2. Are there any topics that you think are critical to add to the body of knowledge?

• yes, following topics should be added: TEI, EAD, GILS, ONIX, CrossRef, MARC DTD, MODS, MARC XML DTD, METS, etc. --Hoangs 09:56, 11 June 2008 (EDT)

• Here again, it's hard to tell what is going to follow (without examining each module in depth, which I am not going to do). So, yes it would be useful to have them exposed to all of the current standards and practices, but this is a three hour class. Why do they only have three-four hours of outside activity? Shouldn't there be more like 10+ hours of reading and activity? The DCMI section seems pretty minimal and dated (see Andy Powell's recent rant on his blog re the Abstract Model). There are more than these three types of interoperability -- what about distributed query? Cross-walking is not a "type" in the same sense as the other two - it is a facilitator of the other two. But here again, I see there is a later section on interoperability.--Schwartzc 09:56, 12 June 2008 (EDT)

• One topic that I do not see covered in your outline is METS. In my intro "Org of Info" course, I introduce MaRC early because it provides a framework for building a single bib record using a variety of different cataloging elements. Similarly, with METS, the students can see how the different classes of metadata fit together: descriptive, admin, preservation, and structural into a coherent whole. This leads me to one other topic that I cover, and it's related to structural metadata: I distinguish between simple and compound digital objects. Thus, from line 15 on page 2, there are cases in which a single metadata record represents multiple individual objects when those objects are tied together into a compound object. Finally, even though there is an entire module on preservation, I would still take the time to distinguish the primary techniques for digital preservation, especially migration versus emulation (and even compare these techniques to the reduction of digital objects to paper for long term storage purposes). Overall, a nice range of topics in the module (I especially like the fact that you contrast embedded (i.e., header) metadata (for crawler purposes) versus separate records aggregated into searchable (and harvestable) dbs; also, I applaud your use of the abstract model in the context of machine processing ... we should never miss the opportunity to convey the "tyranny" of computing systems in terms of their need for explicit specifications! :-) -- Maccalls 14:40, 12 June 2008 (EDT)

• Another vote for METS (and related standards like MODS). Coverage of foundational technologies like XML will be important for understanding most of these standards, but perhaps you do that elsewhere in the course. Caylessh 08:52, 9 July 2008 (EDT)

3. Are there any topics on that you would remove from the body of knowledge?

• No --Hoangs 11 June 2008 (EDT)

• I would move the SemWeb to later in the course and spend more time on the standards that were glossed over.--Schwartzc 09:56, 12 June 2008 (EDT)

• No -- Maccalls 14:41, 12 June 2008 (EDT)

• I agree with moving out the semweb section. It is certainly related to metadata, and in a digital library context, some of its technologies (notably RDF) are used in repository software like Fedora, so I might still introduce RDF as a metadata encoding mechanism here. Caylessh 08:52, 9 July 2008 (EDT)