When I taught intro to internet classes in the public library, I listed multiple ways to locate web pages of interest: 1) search engine 2) directory 3) known item (i.e., reference from a print resource, ad, friend, or other media). This class had to be superficial by its nature. We also know from Rob Capra’s recent paper in IEEE Computer that information literate folks frequently go directly to analogs of print references to find facts online (like phonebooks – organized listings with multiple access points) instead of doing general web searches. In practice, the use of classified or organized web page collections online is perhaps less frequently studied than general web or structured A&I database searching.

At ASIST, I attended a session on this. Also, Jack Vinson recently pointed to a KMWorld article on this. Cataloging professors agree that everything should be cataloged for access so yeah, there is some resistance from those of us who like to search. Busch made a great point that when browsing an online clothing store, you would rather have a categorized list (say: women’s clothing > tops > sweaters) than an empty search box. Steve Papa in the KMWorld article says, “If you search for a monkey in the jungle, it’s tougher than finding one at the zoo, and if you search for unstructured content, it’s tougher than finding structured content.”

OTOH, there is this idea that binge organizing only helps you lose things and creates angst, any imposed structure will have inherent biases, that for a single set of resources there are multiple competing schemes that are valid for particular uses/users (IOW there is no one right classification unless you’re in school). There’s also the idea that search engines are so good that the cost of organizing information is less and less justified.

To add more complication… user tagging is …what? Non-structured if not at all controlled (or faceted) and structured if it’s controlled? Always structured? Doesn’t belong in this conversation (none of the above)?

There’s also structure on multiple levels – whether the data is in a database or free text OR whether it’s indexed or just flowing (so like blogs are structured – there are fields, etc, – ­on one level, and can be unstructured on the other) (see more in the Papa article).

Continuing to wander, it could be that structured information helps the user – even if the user doesn’t explicitly use the structure. For example, in Engineering Village (not affiliated, yadda, yadda), you can throw a google-like search in the easy box – it stems automatically (using a lookup structure, I guess) and it suggests all kinds of terms, codes, fields that may help you find more (or less, but more relevant). EV provides prompts in the latter case to move the user from unstructured searching to structured searching. Is there an example where the structure remains latent yet assists the user? Hmm… not off the top of my head but I may come up with one.

In the ASIST session, Hur-Li Lee reported the results of a study with US and Taiwanese students in which she asked them to find a certain number of professional societies in microbiology. As I recall, her results indicated that an understanding of the field made the students more efficient, because they didn’t go down the wrong path following the hierarchy to the right location.

So, do I have a point? Not really, lol. It depends on the user, multiple methods are still justified. The cost of structure is still justified. More work needs to be done.

¶ 12:49 PM| |cites (technorati) |