Search isn’t relevant, info management is

by

I had a thought a while back about what librarians teach and how they teach it. For the most part it’s “searching”, and typically includes lots of information about how to refine your hits and get better results. This is probably standard practice for most people who teach library courses — myself included. The only problem with this approach is that I find myself increasingly not refining searches, and actually wanting the noise caused by this.

If I were looking for all of the English studies masters theses in our library system, I could provide the command-line instructions to retrieve these, I could do the same in our OPAC with CQL. The thing is that this kind of specialist knowledge is not the kind of thing that there is any real point to trying to teach; if people need to do this kind of thing — and let’s face it: who does? — then ask a librarian. In the specific instance I’ve chosen, the search would involve restricting the search to the local database and searching for a local classification number with a wildcard. This sounds nice and easy, but getting to this solution involves a lot of presupposed knowledge, namely that you know that local classification for masters theses are delimited by subject and are serialized. It also assumes that you’re smart enough to remember that searching for local classification only works when a locale is set. Does a user really need to know any of this?

The user approach would be to set the locale to the local database, search for masters theses (an option in the OPAC) with the keyword “English”. This would return enough hits; actually it would return too many hits. The user approach would be to skim through the hits to find what they were after. Sure, if there are a few hundred hits, then the user might give up — if the information is important enough they will persevere. They might find exactly what they want, or they might find something that is more interesting. Telling the user how to trick the OPAC into showing 1000 hits at a time instead of ten or a hundred is probably a good idea at this juncture.

This kind of reality is where I find myself; I trawl through masses of information every day; not because I’m too lazy to learn how to search properly, but because I find that my ability to skim and evaluate large quantities of data are more effective tools than clumsy boolean searching. I think that this is the case for many young people; search isn’t relevant. The point is that users are getting more and more used to filtering, they’re getting more and more used to being presented quasi rubbish with their search results.

The difference is that I don’t do the searches manually, I’m fed the data and I use tools to help me get to the core of the matter before I even start filtering stuff out. The tools I use, like Bloglines, I use in association with feed searches, technorati, del.icio.us, etc., and these I simply skim every day. Users need to know about these tools and how to make their work more effective — they have the key skills, but they don’t necessarily know about the tools.

The big idea is that information management is where the game is at; I see this everywhere, I see publishers publishing catalogues, but not RSS feeds for new literature, and thus not getting customers. The publishers do provide email alerts, but this isn’t exactly going to ring the bell of anyone who is under forty. If the data isn’t easily transformable into a format where it can be compressed and viewed in the way the user wants it presented, it gets forgotten. I rarely visit the sites I’m interested in, I read most stuff syndicated. I appreciate the work done by recommenders and stream-of-news sites (bibliobabl in Norway), and this saves me having to set up custom searches for this kind of information. I look forward to being able to work with a proper implementation of APML.

All of this isn’t to say that searching is irrelevant, but most of the hardcore stuff will remain — as it always has been — well within the domain of the librarian. This isn’t a bad thing, and it affirms the usefulness of a much-maligned profession. Actually, one notable exception to the “search-for-nothing” paradigm springs to mind, and that is source code. Working with source code requires search skills, and these need to be honed. The fact that tools like Google’s Codesearch implement (albeit rudimentary) support for regular expressions is testament to this. On the other hand, this is again special interest stuff that appeals only to the hardcore user (librarian); for the average mortal there’s the information management concept.

Tags: ,

2 Responses to “Search isn’t relevant, info management is”

  1. Information Overload? « Infonatives Says:

    […] Infonatives Information culture, library weblog for English, Linguistics and Religion @ NTNU (plus some techie/librarian stuff) « Search isn’t relevant, info management is […]

  2. Henry Lewkowicz Says:

    Aside of the methods, tools and policies we use for managing information, we most commonly use searches. Search is one of the critical technologies in finding information but the price for using search is information overload since the search results contain too many choices. A practical addition to any search results or taxonomy browsing would be adding summaries permitting at a glance to see the essence of the text without the clutter of details.

    I’m working with summarization technology and when I do Google searches, the results are automatically summarized and presented according to the keywords of my interest. Such search + summarization approach lets the user much quicker find out relevant documents.

    In fact summarization is a good filtering technology that can be a real asset in reducing information overload while taking better advantage of search engines and any other document management systems.

    If you are interested to try Context Organizer you can find it at http://www.contextdiscovery.com

Comments are closed.


%d bloggers like this: