Nancy Davis Kho Enterprise Searcher Summit West
Jinfo Blog

19th November 2009

By Nancy Davis Kho

Item

This week is the big trifecta trade show for Information Today - the company's KMWorld, Enterprise Search Summit West, and StreamingMedia West shows are co-located at the San Jose Convention Center. I dropped in for the day on Tuesday to hear about innovations in enterprise search technology, but the show's open format enabled me to hopscotch between KMWorld sessions on Enterprise 2.0 and the Enterprise Search sessions. In a keynote for both shows delivered by Andrew McAfee, researcher and author of 'Enterprise 2.0', mentioned a statistic that I heard repeated in every session I attended - the average web search is just 1.7 words in length. (Here's where I confirmed the stat, in a Powerpoint presentation by Prof. Dr. Dirk Lewandowski of the University of Applied Sciences Hamburg - http://digbig.com/5baqte). The challenge on search providers is to find ways to present content in context so that, based on only 1.7 words, meaningful answers are still provided. To me it certainly underlines the importance of organizing the underlying data structure in a way to facilitate serving up the 'right' answers to those abbreviated searches - and who better to do that than librarians? Apparently the folks in the American Office of Nuclear Energy agree. In a session by Wayne Simpson, Idaho National Laboratory Software Architect for the Idaho National Laboratory - Battelle Energy Alliance, he talked about the efforts made to develop a bespoke federated search engine called 'Needle.' It's an open-source-based tool that the INL uses internally for researching across a wide variety of information repositories. Needle has a flexible search interface that allows end users to point at any available data source. A user can select multiple sources such as commercial databases (Web of Science, Engineering Index), external resources (WorldCat, Google Scholar), and internal corporate resources (email, document management system, library collections) in a single interface with one search query. Talking about the multi-year, multi-faceted development process, Simpson stopped and said, 'You know, if you're working on federated search and don't have a librarian on the team, you're making a mistake.' He went on to say that the librarians were instrumental to getting the Idaho Lab's federated search product to where it is today - launched internally, and with plans for a public rollout in the future. It was as good a plug for the science of librarianship as I've heard - and a good reminder for the ESS audience.

« Blog