Making Sense of Search
By Marydee Ojala
Editor • ONLINE
Everywhere I went this past summer, whether physically to library and information professional conferences, such as SLA and ALA, or electronically to information guru blogs, I kept hearing the same refrain: Search is no longer important. As both a word and an action, it’s been diminished, downsized, devalued, and disregarded. Dissing search seems to be the “in” thing to do. Apparently the substitute terms are findability, discoverability, or, simply, answers. Has search really become irrelevant? How do we, as information professionals, make sense of search?
Findability, a word popularized by Peter Morville, a contributor to this magazine in past issues, strikes me as the counterpoint to search. If you architect a site to be findable, then the search function will be successful. Discoverability is different. It implies a degree of surprise, of the unexpected. Text mining (Ernest Perez reviews a text mining product in this issue) is a form of discoverability, as is the power of serendipity and of pattern recognition. I used “discoverability” in my column in this issue to describe what could be discovered that I didn’t expect to uncover through normal search mechanisms.
The problem with search, it appears, is that it doesn’t necessarily find the desired answer. This could be an information architecture problem, a search engine algorithm hiccup, or a lack of query understanding. Entering words into a Web search engine, or into an enterprise search engine, doesn’t guarantee a result that matches what the requestor had in mind. Conference speakers talked about intuitive searching, holistic analysis, finding conversations, and content management.
Clemens Ceipek, of LexisNexis Corporate Markets, suggested that searchers want a full-fledged answer rather than a list of documents. “It’s not about searching,” he added, meaning that information professionals need more than the proprietary, licensed data typically viewed as LexisNexis’ purview—as is the case with the other premium content companies. He’s probably correct, and I applaud the efforts of LexisNexis to move beyond documents to add blogs, wikis, and unstructured internal information. LexisNexis is not alone in this; other companies have realized the same thing and are rushing to add 2.0 capabilities and information to their offerings.
I don’t believe that search is irrelevant. It’s not even completely broken, although it is sometimes misapplied. Algorithms that work for an open Web search do not necessarily perform well when applied to internal, unstructured data. What’s changed about search is our expectations. Search is now supposed to deliver accurate, analyzed answers. The onus is on the search engine to perform rather than the searcher to perform. Search is dismissed because it doesn’t anticipate the searcher’s every need. Web search engines now deliver good results for simple requests. However, to make sense of search requires human intervention. It’s the information professional who takes search results and analyzes them, bringing to bear human intuition and subject knowledge. Information professionals provide the link between search and analysis. Our expertise enhances search, incorporates findability and discoverability, and makes search relevant.
is the editor of ONLINE. Comments? E-mail letters
to the editor to