Wednesday, March 28, 2007

Another thing that makes me hungry: del.icio.us

Perhaps I am not eating right -- just the mention of del.icio.us makes me hungry. Who knows, that may be part of my problem with the site in general; the word delicious for me has a lot of power, and promises something that the site (for me) doesn't deliver.

I think it's key that del.icio.us is seen as a social bookmarking tool. Making your bookmarks/favorites portable is an old need (and really, most hompages in 1994 were in nature not unlike a user's del.icio.us page). But it's a need that can be filled with numerous tools. I suppose the greatest value of del.icio.us is that it shares out those bookmarks and the tags that go with them. When scaled up to the size of the Internet, that becomes powerful. I am not convinced that it regularly outperforms Google in general research. An example we're in here lately is looking for primary source material from pre-1900 eras. Looking in del.icio.us for tags {primary source} and {primary sources} gives useful results (although, it should be clear, different results due to the little "s" -- which makes anyone who's used library catalogs with LCSH shudder a bit). But really, Google gives the same results. To sum up the practice search (actually, a real user question from 3/27):

in del.icio.us
search (in tags) for Primary Source Augustus: no results relevant to the Roman emperor
search (in tags) for Primary Sources Augustus: a good source http://virgil.org/augustus/primary-sources.htm on the Roman Emperor comes to the top, and that's it

in google
search for Primary Source Augustus, or Primary Sources Augustus and the good source at http://virgil.org/augustus/primary-sources.htm comes to the top. Plus lots more.

So I suppose del.icio.us is most useful for depth on topics of interest, and, of course, for those web 2.0 stalwarts, discovering RSS feeds and linking up with users of like interest. With users I'd still probably be more inclined to use Google... the nature of the Google search engine is that it's going to pick up on things that are tagged frequently in del.icio.us unfailingly.

A note about tags: at least they give many users a language with which we can explain how MARC fields allow field searching. But the lack of rules for applying tags make them only so useful, especially when not scaled to net-wide implementation. I think I'd be excited it Horizon could implement user tagging for bib records that would then be used to facilitate new, user-centered search options for the HIP -- but unless we were sharing the tag database with the entire customer base of Horizon (and perhaps ibistro as well), the small scale of taggers would make that data completely idiosyncratic. Probably pretty interesting, but probably not that useful for information retrieval.

Tagging is to cataloging as humming a half-remembered jingle is to symphonies
but
Tags are to cataloging as (not-by-numbers) painting is to photography

{i will write metaphors for food. if it's del.icio.us}

No comments: