Category: Enterprise search & search technology (Page 44 of 59)

I’ve been thinking a lot about the best models for enterprise search lately because I work with clients who are mostly unhappy with the way their current enterprise search technology doesn’t result in meaningful content results, or because they are trying to find better ways of categorizing the content for easier access. One technique that I use to elicit feedback on possible approaches is to find examples on the Web of search interfaces that I believe are worth consideration. While I work with corporations with a heavy amount of R&D related content, I use examples as diverse as UN sites, catalogs, health care organizations, and so on, to get everyone’s imaginations working on the possible ways we could present content search options.

In some cases I am working to achieve a browsable construct for a taxonomy (that doesn’t necessarily look like a conventional taxonomy) and in other cases I may be trying to expose the searcher to “advanced” search features without getting into explanations of Boolean options, while still supporting them.

I have recently found a mental digression by thinking more about the reactions I get when I forward links to my clients for “design consideration and feedback.” The reactions seem to be quite visceral and, I’ll admit, mine are, too. I am beginning to segregate likes and dislikes into highly textual interfaces with very sparse graphics vs. interfaces that offer (or attempt to offer) a highly graphical layout of the window. Personally, I have no problem with graphics when they fit or mesh with the text but I realize that I ignore most pictorial graphics. Even when I attempt to use symbolic icons in a graphical interface I encounter for the first time, the struggle to connect meaning to the picture is not worth my effort.

The most confounding interfaces are those with a lot of text and a lot of pictures all mixed in, especially without a cohesive and minimalist color palette. I remember a strange disconnect several months into using Google. A significant holiday day came when they jazzed up their Google imprint. I was certain that it reflected a change in product design and “I didn’t like it.” When someone assured me that it was just a little “Google” fun, I accepted it but I still don’t like having them mess with the pure interface. When they moved the “directories” tab from the main page, it annoyed me and I don’t use it nearly as much any more, first because it is on a new page and second because it has a little picture attached that doesn’t mean “directories” to me.

Guess I’m still mired in the IBM “KISS” mode but I do like my text clean and simple. Take a look at Siderean’s demo – just the way I like it, no frills. No pictures are worth a thousand words to me.

There is an enlightening discussion going on between Lou Rosenfeld, Clay Shirky and others on the utility of folksonomies as used by Flickr and del.icio.us, vs. subject-matter-expert developed taxonomies. As one of the commenters has pointed out, this is not an “either/or” issue. Certain applications where the scope of the content and users is bounded will benefit from the discipline of a carefully architected vocabulary. Other applications where the scope of either the content or the user community is less well-defined will either suffer or, more likely, the users will ignore the prescriptions (this is why the “semantic web”, if I understand it at all, is hopeless). The key issues are related: cost and adoption (cost is usually a function of adoption, not development), and I think they both would agree on this point. How these approaches might work together is trickier and well worth exploring. In any case, this debate provides a condensed lesson in many issues that most enterprise content managers have probably not thought through, but even those that have should check out this thread.