The second engine that I used was the Healthgate MEDLINE Search. This method had some definite advantages over the regular PUBMED search engine. The main one was that it would allow for advanced searches, rather than just the simple ones that the PUBMED engine allows. Searches can be limited by date or subject matter, or by where in the documents the keywords are located. I really did not take advantage of all of these options, as I simply searched for “lung cancer” and “treatment,” both in all fields for the first question, and then “knee injuries” and “sports” in all fields for the second. Allowing a patron to limit a search to articles pertaining to humans is also a useful feature, as is the option to only show results published in English. The other searches resulted in several articles in foreign languages. Although the abstracts were in English, the foreign-language articles would probably not have been particularly useful to the imaginary users in the situations. The results seemed slightly better than those for the first search engine, with more relevant articles located closer to the top of the list. If I had taken full advantage of the advanced searching options, the results might have been even better.

The third search engine that I used was ConceptLink, which was very confusing for me. Its search engine was not particularly user-friendly. It required that a user type in one term that appeared in its database, and it would provide a list of related topics. This system could be useful, but its main problem is that a user must hit upon a term that the search engine programmers had used. Most of the terms that I tried, including the fairly basic “lung cancer,” produced no results. When I finally did receive some hits, a complicated process involving applets was necessary to add additional terms to the search. With Healthgate, the terms could be combined immediately, by using Boolean operators. Because of the complicated search system, ConceptLink retrieved more articles that were not strictly relevant to the topic than the other two engines did, and I knew of no way to remove the foreign-language articles from the results. If someone gained some experience working with this engine, it might prove useful, but it is not a good tool for beginners.

Search engines certainly have a place in future digital libraries. Computers are becoming increasingly important even in physical libraries, and they are an integral part of digital ones. In order for a search engine for a digital library to be truly effective, it should be easy to use, but also allow for more complex searches. Giving several search options, with the simplest one as the default, would probably be the best way in which to accomplish this. A complicated system like ConceptLink could be incorporated into the digital library’s search engine, but it should not be the only method of searching, as that would intimidate beginning users, and not provide the most accurate results possible.

Return to my main INFO 653 page. 1