Issue of the Month: September 2007Federated Searching
Jennie Burroughs, Assistant Professor
Government Documents Librarian
In August, the Mansfield Library launched a new service: 360 Search. You can access this service if you click on the “Databases: Search by Subject” or “Databases: A-Z” links on the library home page. 360 Search is an example of federated searching.
What is a “federated search?”
Federated searching has also been referred to as “metasearching.” Essentially a federated search strives to be a one-stop shop that allows users to search many information resources simultaneously: library catalogs, article databases, and websites. The results of a federated search like 360 Search are a mix of books, articles, government reports, webpages, and other information sources. Google is working toward a federated search-type idea they refer to as “universal search” that will compile items from Google’s various search engines (Web search, Blog Search, Google Scholar, Google Books, etc.) into a single results list. However, 360 Search offers something that Google does not. As helpful as they are, general search engines like Google and Yahoo! do not search the “deep web” (also called the “invisible web” or “hidden web”), which includes library catalogs and much of the peer-reviewed, “premium” content in databases the library subscribes to for university researchers (Fryer, 17). 360 Search is a quick way to get access to that hidden, scholarly content.
How can federated search help researchers?
Librarians often hear from users who would like to search all of the library’s materials with one simple search. Federated search is the first step in that direction. 360 Search reduces the number of decisions you need to make at the beginning of a search (Baer, 518). Instead of figuring out which database to use first, searchers can try out many databases at once and see which ones have the best resources on their topics. Being able to use multiple sources at once speeds up the search process as it reduces the need to rerun searches a database at a time. For expert searchers, a quick federated search may reveal unexpected articles in sources they hadn’t previously tried or considered. Also, the citation export feature for 360 Search is quite smooth, so researchers that use RefWorks or EndNote can use the federated search engine to quickly grab citations for their preferred bibliographic management system.
How does it work?
Federated search engines are much more complex than they seem. In a nutshell, here’s how federated search works with library or subscription-based content. The library first selects and pays for a federated search service, and then librarians select the most relevant, high-quality sources for each subject. These subject clusters of databases, catalogs, and websites provide a focused search and enhance relevancy for results. When a searcher enters keywords and clicks on “search,” the query goes to the vendor of the federated search service, the service automatically runs multiple searches in the designated sources simultaneously, and compiles the results. The service then de-dupes the records, automatically generates subject clusters to help the user define relevance, and creates links to the full text (when available).
What are the drawbacks to federated searching?
Even those who see promise with federated search engines note that they are a “great starting point, but never the ending point” (Fryer, 19). Federated searching does not allow for particularly precise searching. This is because the information architecture of information sources can vary greatly. Databases may use different fields, select different symbols for wildcards and truncation, use different subject headings and descriptors, or may disallow Boolean searching–all of which make it difficult for a single search to be run the same way in multiple sources. A federated search will often be slower than a search in a single database or on Google. Deduping does not always work perfectly, and searchers will sometimes see results that don’t seem relevant, especially if they are searching outside the context of a specific subject (for instance, a search for “magellan” could refer to the explorer Magellan, the space craft Magellan, or the mutual fund Magellan). Expert searchers and those doing intensive research may prefer using a database’s original interface. Searching directly in a specific database will provide greater control over searches and offer more-refined results and additional features.
Federated search products are still in their early stages. Vendors and libraries are making improvements to these services all the time, and the Mansfield Library will continue to adjust the 360 Search service to meet the needs of the University of Montana community. If you have feedback about the new 360 Search, please let us know.
For additional information on federated searching, try these sources:
Baer, W. (October 2004). “Federated searching: Friend or foe?” C&RL News. 518-519. (Accessible to UM students, staff, and faculty)
Federated search. Wikipedia. Retrieved August 14, 2007.
Fryer, D. (March/April 2004). “Federated search engines.” Online. 28.2. 16-19. (Accessible to UM students, staff, and faculty)
Jacsó, P. (October 2004). “Thoughts about federated searching.” Information Today. 21.9, 17-20. (Accessible to UM students, staff, and faculty)
Marshall, P., Herman, S., & Rajan, S. (2006). “In search of more meaningful search.” Serials Review. 32.3, 172-180. (Accessible to UM students, staff, and faculty)
Surratt, B. “Federated search engines: 2001-2003.” Association for Library Collections & Technical Services (ALCTS). Retrieved August 14, 2007.
Tennant, R. (June 15, 2003). “The right solution: Federated search tools.” Library Journal. 128.11. 28, 30. (Accessible to UM students, staff, and faculty)