Federated Search Exploration

What do you think?

Give us your feedback! What do you think of federated search? Would this be helpful or harmful to our users? What kind of resources would be useful to be able to search simultaneously?

You can jot down thoughts here, or use the discussion tab above!
Here's a video tutorial on contributing to the Wiki - it's easy!
See Katie, Matt or Tammy for help!

Type your comments here!

Email from Fran:

Here’s a few links to some multi-search products I found at other market-basket libraries and I think are linked on your wiki. The Xerxes product looks interesting as perhaps a first implementation.

Rochester Institute of Technology “Library Database Portal”

Georgia Tech Library’s “Article Cross-Search”

MIT Libraries’ “Vera Multi-Search”

Xerxes Used by GT; RIT and other libraries:

Notes from group discussion (These were kind of long, so I put them on their own page. -KD)
Recording of presentation / discussion
Here's a link to the student audio that Matt plays during the presentation: http://www.rpi.edu/~benzim/fedsearch/googleIt.wmv
KD chatted with Michele: KD's notes

Federated search is a good idea. Seems like it would be particularly useful for people working outside of their majors/disciplinary areas, where they aren't familiar with the databases. (Like a management student looking for engineering information.) Hard to tell what database to use out of the big list of databases. It would make things easier for users, especially if it included SFX links.

MM said that doing library research in the classes she took was easy because of her experience working reference. She knew what databases were available and where to find information. But people in her classes that didn't have her experience had no idea where to start.
KD chatted with Mary Dixey: KD's notes

- Mary feels like the technology involved in federated search, discovery interfaces, etc. is still changing rapidly. Things work way better than they did 5 years ago, but are still improving / in an incubation period. Probably this is partially due to more competition - there are more options out there now.
- Philosophically speaking, this is the direction we need to go in.
- People expect something like Google - single search box, relevance ranking.
- Saw EBSCO's discovery interface at ALA Midwinter - looked pretty good.
- Would be nice to be able to toggle whether the catalog is included or not, for people who are off-campus / distance ed students. It would be nice to be able to limit to resources that are available to them where they are.
- Difficulty of searching different kinds of resources together - vendors are struggling with how to address this within their own resources - how do you search company information, etc. along with journal articles, in the same interface?


Bob stopped by to say that he is in favor of moving away from silos of information and toward a single search box, whether through federated search, discovery interfaces, or whatever. "We have 1.8 million dollars invested in electronic resources, if people can find them." (BM via KD)


KD chatted with Linda Knaack: KD's notes.
- She was on a big regional committee looking at federated search a few (4?) years back. (Central and Western Mass Regional Library System) - public libraries and academic libraries - mixed bag. Compiled a report for the exec. director of CWMARS. The group didn't go ahead with federated search - a few subgroups were thinking of looking further into it, but she doesn't think they did.
- Linda is concerned about the resources that aren't included, and not being able to distinguish between the resources, and cost. At the time she was on that committee, there was a big range of prices. Was helpful to actually be able to see the stuff in operation - lots of people were on that group, so lots of good q's were asked. Christine Haggstrom ran the group - the process worked well. She works at Innovative now (maybe track down her and ask her for ideas on how to manage this project if we continue to investigate) Created tables with comparisons of features for the different products, and a final report with recommendations.
- Linda thinks that at Hartford, with 500 fte or so, and way fewer resources, such a technology isn't as vitally important. biggest complaint she encounters from students is that the databases are awkward to search - not so much that they have to search in multiple databases, but that the individual databases are hard to search and not like google, which is what they're used to.
- She has another part-time job at a community college. Those students think databases are hard to figure out, too - the interfaces. Google is easy for them.


Unfortunately I was unable to attend the presentation, but have a strong interest in the topic. I would echo other comments that although the technology is not perfect, we are losing our patrons to Google and must offer a simpler interface. This does not mean we don't education our users to the more sophisticated tools we do offer, but let's not lose our user base. I would also agree with Tammy that there are big staffing impacts with every new technology and we are pushing our limits. A conundrum. (PH)


John stopped by to reiterate his point from the discussion about the importance of having a tool that is useful to both undergraduates and seasoned researchers. Some things that might be useful to more experienced users are the ability to search subgroupings of databases that are relevant to a particular discipline and the ability to limit by format. -KD


Here are some thoughts from my student assistant, Michael, who sat in on the presentation last week (TAG):

"I think pursuing the federated search option may create a new bridge between the virtual data of the internet and resources in the library, the latter of which has been partly neglected with the advancement of internet technologies. Although the system may have some drawbacks, i.e. the real time search, I think they pose unique opportunities for the library to be innovative in solving other problems.

The loading screen possesses a lot of potential; not only can it relay information in real time about the search query, but it can also exist as a vehicle to display randomized useful tips or information about the library. These could range from short updates about library events, a brief explanation of the dewey decimal system or the library of congress, to information on using the inter-library loan system. Over time and with multple search queries, users will learn a variety of information that can help them in current or future situations.

Transparency is another quality of the federated search I appreciate. Even for experienced researchers, this kind of transparency can certainly open doors or present new opportunities for people to work between fields. Although someone may be very knowledgable or experienced in researching journals specializing in biology, there could be information in journals of other fields that hold key information. In this sense it works in a very interdisciplinary dimension, something I think the Institute strives to do." (MV)


I think federated search and discovery layers are very promising developments. They can increase efficiency when searching across multiple systems, identify unanticipated resources for both inexperienced and seasoned researchers, and address evolving user expectations. However, we must consider the impact of continually adding new products and services. Each new system requires some level of management and maintenance, and we need to find ways to support new technologies without placing an undue burden on our systems staff. (TAG)