Searching the Web: Toward Maximizing Relevance

  1. Mark A. Spasser, PhD, MLS
  1. Jewish Hospital College of Nursing and Allied Health, St. Louis, Missouri
  1. REPRINT REQUESTS:
    Mark A. Spasser, PhD, MLS, Chief, Library and Information, Services/Associate Professor Jewish Hospital College of Nursing and Allied Health, 306 S. Kingshighway Blvd., MS:90-30-625, St. Louis, MO 63110-1091, Telephone: 314-454-8171, Email: mas1200{at}bjc.org

Searching for information on the Web is paradoxically easy but frustrating. There is an increasingly vast amount of high-quality information available on the Internet, but finding it can seem like groping through the world’s largest haystack for the proverbial needle. This review discusses some of the most basic and important methods and techniques, (table 1) applicable to almost any search using almost any search engine, whose thoughtful application ensures that the information found is both high-quality and relevant.

Table 1.

Useful search features common to most search engines.

TYPES OF SEARCH ENGINES

Centralized - Google, Teoma, and AlltheWeb

Specialized/expert - Pubmed (Medline), AllLaw, MedHunt

Meta search engines - Meta engines are becoming less important due to centralized search engines indexing the Web so comprehensively, and as a result will not be covered in this review

GUIDING PRINCIPLES

Be specific: Centralized search engines, such as Google and AlltheWeb, work better when you use multiple keywords that are unlikely to pop up on irrelevant pages. Also, of course, use specialized search engines whenever possible.

Evaluate retrieval: Listing length of candidate documents returned by queries submitted to search engines is directly related to amount of evaluation time needed.

Gauge time: Whereas actual search time in getting candidate listings from search engines is relatively fast, document-by-document download and review is the time consuming part of the process.

Read instructions: Analogous to database interfaces, every search engine has its own syntactical peculiarities; entering a query properly can mean the difference between finding what you’re after and being deluged by irrelevant links.

Go Boolean: Most sites permit the construction of queries using the AND, OR, and NOT Boolean operators (use parentheses if you use more than one Boolean term). Also the + and − implicit operators, are almost universally supported.

Be exacting: Surround words with quotation marks to search them as a phrase. The utility of this widely supported search feature to ferret out only the most relevant information can’t be overstated.

Get advanced: Many engines come with an advanced search page offering additional query types and sorting options, which are especially helpful when a simple search bombards you with pages of links.

Be steadfast: If the first site is not successful, try another; bookmark several and try new contenders from time to time. Each engine has potentially complementary strengths and limitations.

Save pages: Once documents are located, keep track of them; at the very least bookmark them.

CONCLUSION

Time is well-spent in learning to pose a proper query and to take advantage of the way search engines work.

Search engines mentioned in the review:

Google: http://www.google.com

AlltheWeb: http://www.alltheweb.com

Teoma: http://www.teoma.com

Ditto: http://www.ditto.com

AllLaw: http://www.alllaw.com

MedHunt: http://www.hon.ch/MedHunt

  • Received August 1, 2002.
  • Accepted August 13, 2002.
| Table of Contents