Federated Search Journey (some answers for Sol)

Posted by Chris Carmichael on January 23, 2009 in Federated Search, project management, Search |

The Federated Search Blog was kind enough to pose some questions to me about specific aspects of our “search for federated search.”   So, to keep the conversation going, I will answer some of them here.

Q1. How did you decide which sources to add to your WebFeat implementation?

A1. The reference department (a team of 7 people) went through a somewhat arduous process, but only because as the team leader, I wasn’t clear about exactly what I wanted all of us to consider.  I originally asked each person to give me a list of their top 15 databases that they would like to see federated.  That’s it – no further guidance, no discussion of what they should consider, zilch.  Four days later it became evident that I needed to provide some context.  “If you were going to search multiple databases in a General and News capacity, which ones would you choose?”  “In your subject areas of expertise, which databases would you want to federate?”  “Are there specific resources you WOULD NOT include?  Why?”

It then became clear that selecting a “Top 15 List” was pretty useless.  If we’ve got 50 slots, let’s pick which ones we each would want and then tally which ones got the most votes.  We agreed on a couple of things up front:

1) The sources we federated had to be ones we paid for (for now).

2) Sources that had a limited number of simultaneous users or that ran as a pay-per-search entity were automatically eliminated

Q2. How easy has it been to configure your implementation?

A2.  Well, it’s never as easy as it looks during a vendor’s presentation because each client’s technology environment is just a little bit different.  That said, the administrative module interface is nicely laid out, it makes logical sense in terms of what order particular tasks are done.  My own issue currently is how the service will act with our EZProxy environment.  I do have to say that the tech support/customer service shop has been extremely responsive and they don’t make me feel like an idiot.  (I swear, technology can sometimes make even the brightest people feel like they know absolutely nothing.)

The other thing that has been nice within the administrative interface is the capability to make categorizations that match our current systems’ language.  We are trying very hard to make our systems, disparate as they are, say the same thing in the same way.  Consistency is key to learning, so if we are going to throw a new discovery tool at people (students, faculty, staff), we need to be very cognizant of what they already expect.  Standardizing verbiage across the catalog, the database interfaces, the website, and the search platform goes a long way toward making people comfortable with new tools.

Next post: more answers!  –ctc


Copyright © 2007-2022 Carmichael's Connundrum's All rights reserved.
Desk Mess Mirrored v1.7 theme from BuyNowShop.com.