6 Responses to “Podcast is back on the air!”

1. Rob Wall says:

   February 4, 2005 at 9:02 pm

   Yay! I’m downloading it right now.

   Reply
2. Matt Pasiewicz says:

   February 7, 2005 at 4:41 pm

   why not supplement taxonomies with folksonomies and combine the best of both on a massive scale. Using info professionals to review and help link and/or normalize folksonomies. Are the two incompatible? It doesn’t seem like they are too me. There has to be someway to bring the best of both together.

   Also, another thought on taxonomies … even the best of librarians might miss an outstanding way to classify an item … sometimes deep domain expertise could be just as interesting and valuable when classifying content (even better when supplemented with the value of an information professional) … this is where the massive scale of folksonomies, combined with some more professional review and connection to other items in a formal ontology/taxonomy, might be really interesting.

   BTW, the project you mentioned sounds really interesting … sort of an automatic classification and metadata entry engine with human confirmation. Nice. I’d love to learn more!!!! We want to do something really similar.

   Reply
3. D'Arcy says:

   February 7, 2005 at 4:48 pm

   Matt, exactly! The two different types of classifying a resource should be completely compatible! It just depends on the audience you’re talking to… Talking to the Average Joe about taxonomies will likely kill any real contribution from them, but telling them they can just tag whatever they feel appropriate, knowing others will do the same and eventually you’ll cover all bases – well that’s something people can do. And, if librarians are pitching in as well with their more refined metadata skills, then kudos all around.

   Many projects will have many resources that don’t require deep classification, though. Imagine having a librarian tag every slide in a 10,000 slide collection of embryology microsopy slides. While it would be ideal, it would cost a fortune. On the flip side, if you have a small set of very rich resources (websites for teaching chemical reactions, for instance), you would be better off to have the rich metadata from the start so it can fit into curriculums (curricula?) more effectively…

   Reply
4. Matt Pasiewicz says:

   February 10, 2005 at 11:17 pm

   Agreed. It should be a fun few years as we learn how to create/identify what really works and what doesn’t … and to what extent some critical mass is cruical to effective tagging (or maybe not if there are especially good domain experts in a smaller group – i dunno). Thanks for the conversation!

   Matt

   Reply
5. Amy Gahran says:

   February 12, 2005 at 12:54 pm

   D’Arcy — thanks so much for this audio post. I listened to it this morning, and suddenly found myself getting really jazzed about the folksonomy concept for the first time. This set me off on a flurry of research, collecting items that had been on the periphery of my professional awareness and compiling them into a major “grab bag”

   You can check out the results at: http://blog.contentious.com/archives/2005/02/12/labels-and-metadata-grab-bag

   And yes, your podcast is on the top of that meaty list.

   My background is mainly in writing, editing, and journalism. My colleagues in those fields could definitely enhance their work (and career prospects) by learning more about metadata.

   Specifically, I’ve thought for a long time that writers and editors should be more directly involved in metadata. Unfortunately, many of those professionals have a strong aversion to creating and working with metadata — at least, the more structured ontologies. It just feels too much like programming to them, it doesn’t resemble the way they organize information in their heads. However, folksonomies might be a great bridge for these people, since it’s a much more human and organic approach to metadata.

   Thanks again,

   - Amy Gahran
   Editor, CONTENTIOUS

   Reply
6. D'Arcy says:

   February 12, 2005 at 1:11 pm

   Amy, thanks for the feedback (and the collection of links you’ve gathered!) I’ve got no research to back it up, but my gut tells me that people have trouble thinking natively about rich/deep taxonomies. Our brains are excellent at classifying things and organizing things into patterns, but it seems to work best when it’s got an aggregate of items to work with rather than one highly structured piece of metadata.

   Also, why aren’t the software tools doing more of the heavy lifting with respect to classifying content? Full text indices of a document should be able to provide at least a basic description of the content, and a set of relevant keywords automatically… And why wouldn’t it make sense to go one step further and analyze a set of documents for relationships based on content, rather than relying on people sitting down and doing that manually? There’s a lot of work to be done here…

   Reply