One of the welcome recent developments in information delivery is the ability to search across multiple resources simultaneously. This strategy--sometimes called cross-database searching, global searching, or multi-file searching—increases search efficiency and fosters the discovery of unexpected resources.
Other articles in this issue discuss cross-database searching in the Ovid and DIALOG interfaces and NLM's new global search engine for National Center for Biotechnology Information resources.
While the advantages of cross-database searching are evident, users should be aware of some considerations.
Varying levels of sophistication are used in devising cross-database searching. The best systems control for variations in database structure by mapping search terms to their equivalent fields in different databases. For example, the controlled vocabulary terms for different databases might be called descriptors, keywords, or subject headings. Smart cross-file search software will make searching these fields effortless.
The best use of cross-database searching is in identifying quickly which information resources are likely to have information on a given subject. While Medline is generally the best place to find journal articles on medical topics, areas like genomics or psychosocial issues may be captured in a variety of resources. Searching across databases can find troves of articles in unexpected places.
While most of the resources discussed in this issue are free, DIALOG databases are fee-based. However, DIALOG provides a free Internet search that will identify the number of articles on a topic in any of its hundreds of databases. While the individual article titles cannot be seen without paying to search that particular database, the free feature spares users from expensive searches of off-topic databases.
Of the cross-database searches discussed in this issue, the Free Library of Philadelphia’s 1 Search is the only one to search across databases from different producers with a variety of interfaces. It uses WebFeat, a sophisticated commercial software program. Users choose those databases they want to search. Results are listed by the database in which they were found. Choosing a citation will take the user into the original database.
Start a cross-database search with a keyword search of likely terms. But, take advantage of each database’s features and don’t stop at those results.
Some advice:
- Review the number and uniqueness of citations pulled from various databases.
- Consider the value of the articles unique to each resource and determine which resources have good results for your topic.
- Then rerun the search in each of these databases. Follow database cues like subject mapping and examine results for additional search terms that can be used to broaden your search to find similar articles.
- Finally, use the database’s limit features to refine your search.
Related Links:
