Sorry for the delay in posting the rest of this, but we were flying to Florida yesterday to visit family. Nice to be in consistently warm weather for a few days. 🙂
When library catalogs were first developed, all of the electronic needs of a library system were fulfilled by the ILS. At the time, there was the cataloging side, the circulation side, the acquisitions side, and the catalog of the library’s print holdings (the OPAC). Things have changed significantly since that time. We no longer only have a purchased collection. We have electronic collections we lease, collections we create through digitization, and scholarly digital repositories. Some of these collections can cost tremendous sums to lease access to or create. But can they be found within the integrated library system? No! There is no common access point for all of the libraries resources! The more electronic collections that are available, the greater the variety of search interfaces users must contend with. And most users are not going to use eight different interfaces to find eight different journal articles. I was dumb enough to put up with that sort of frustration as an undergrad and in my first graduate school program. I don’t imagine that students who are younger than me and have higher usability expectations are willing to go through such a hassle to find a few resources for a paper.
There are several promising developments that have been developed over the past few years and can help librarians to integrate their holdings, reduce complexity, and make it easier for users to search. The first is Federated Searching. Federated searching allows users to search multiple internal and external holdings of a library. This can include the traditional catalog, subscription databases, and free web resources. With each search, federated search engines queries various databases or catalogs to obtain results from all of them. Just imagine searching for information on a specific subject and getting results that include books at the library and full-text articles available on the Web. The number one benefit of federated search is convenience. It’s just easier
There are several shortcomings of current federated search offerings:
- When content is retrieved from diverse sources, it is nearly impossible to accurately rank the results. Many federated search tools rank the results from each source separately or they do not rank them at all.
- Searching various databases will often net duplicate results as many databases offer some of the same journals.
- Because it is querying a variety of databases, the federated search tool must be simple. It cannot offer a lot of advanced search options because not every database it queries has those same search options. This may not appeal to professors or savvy searchers who like to narrow their search using advanced search options. And ending up with hundreds of unranked results from eight different databases may not work for the student looking for just a few articles.
Some people say that Google Scholar marks the death of federated search because you can easily get results from a variety of databases in a single interface. I think designers will find ways to minimize the flaws that are inherent in today’s federated search engines. But I do wonder if link resolvers and metadata harvesting will make federated search obsolete.
The problem with federated searching is the paradigm. Simultaneous searching in real time is ludicrous. There is a better way, a proven way, and yes, federated searching as it is done now is not viable.