What are the difficult challenges that federated search vendors are facing? How are they meeting these challenges? Miriam Drake, Professor Emerita at the Georgia Institute of Technology, and one of the judges of the federated search writing contest, asks a number of industry leaders some difficult questions. Drake’s article appears in the July/August edition of Searcher Magazine. The article is not freely available; only this short description appears on the publisher’s site for the current edition:
Miriam Drake talked to a number of federated search developers to get their perspective on issues such as scalability, common names, and reliability and how these areas affect federated search technology and user satisfaction.
While the article could be misconstrued to be yet another rendition of “what’s wrong with federated search,” this piece is different. Drake isn’t interviewing the librarians or the users about the problems; she’s unabashedly going straight to the application designers and probing them on what they’re doing to satisfy the users who complain that federated search isn’t as easy as Google.
The article is most interesting in its quoting of the industry experts. Asking the federated search developers for a basic definition of federated search elicits fundamentally different responses from two respondents, not so much on what the technology is but on its major purpose. If I were shopping for a federated search solution I’d be very interested in what the people who are influencing the industry think the point of the technology is. Wouldn’t you?
Drake asks about scalability: what are the problems and how are vendors addressing them? Responses are diverse and include discussions of the semantic web, taxonomies, and source selection algorithms. Again, it’s interesting to read what each developer is most concerned about, in this case, regarding scalability.
Ask developers about presentation of search results and, again, you get a different focus from each of them. While clustering of results is a common goal, each vendor puts a different spin on what’s important in clustering. Randy Marcinko, CEO of Groxis, had a most interesting perspective on clustering. Drake explains:
Marcinko prefers the term “lenses” rather than clusters. Lenses can be adjusted to reveal a panorama or zoom in for detail.
Drake’s article explores a number of areas: main purpose of the technology, scalability, name searching, user profiles, alert services, presentation of results, multi-lingual searching, user training, and library OPACs and concludes with these thoughts:
Federated search systems could be improved with central indexing, taxonomies, visual as well as text representations of results, foreign language translation, and user training. … Making a search system easy and effective for users requires a huge amount of behind-the-scenes work and cost. My conversation with developers and service providers convinced me that they are working hard to make their systems better.
Miriam Drake interviewed the following persons (in order of “appearance”):
- Raul Valdes-Perez, CEO of Vivisimo
- Stephen DiStasio, product manager at Serials Solutions
- Stephen Abram, vice president of innovation at SirsiDynix and current president of the Special Libraries Association
- Randy Marcinko, CEO of Groxis
- Abe Lederman, founder and president of Deep Web Technologies
- Christy Confetti Higgins of Sun Learning Services at Sun Microsystems
- Barclay Hill, Intel librarian