CIL2006, Day 1: Federated Search Engines – Lessons Learned

Frank Cervone:

70% of all searches are keyword searches and they pretty much get article searches

Students don’t understand the concept of “metasearch” and federated

Federated product – good place to go for starters

students have strong expectations about how results should be displayed: relevance order – it’s the search engine model…

advanced users tend to head towards databses

If it’s not federated, it’s ignored. The hope is that people will click through to the native interface when appropriate

finding the right group of databases for subject areas is important

long lists of databases – students find them confusing and make them feel stupid

they group databases by “best bets” or the three major databases in any given topic area.

It’s critical that they work from the perspective of the patron.


Jeff Wisniewski:

Webfeat – live since Sept 2004

majority of searches come from the quick find search on their website’s main page

They provide three access points: federated product, a-z list, and subject list

Google has set the speed standard – they get “it’s kinda slow” comments

Speed constraints – be selective – dont’ want a “earch all” when all equals 300+ databases

monitor usage stats, especially turnaways

implement a formal evaluation process


Ying Zhang: MetaLib Implementations

spoke about her organization’s implementation of MetaLib


Athena Hoeppner:

Usability aspects of their federated search product

Most users use the quick search feature

They believe customization would help – trying to label things differently

Metalib isn’t an ideal solution for them because of the lack of easy customization

They’d like to add lots of help features, add useful icons, and have the visual design mirror their website