Regardless of the search tool being used, the development of an effective search strategy is essential if one hopes to obtain satisfactory results. A simplified, generic search strategy might consist of the following steps:
This strategy should be applied to a search of any electronic information tool, including library catalogues and CD-ROM databases. However, a well-planned search strategy is of especially great importance when the database under consideration is one as large and amorphous as the World Wide Web. Along with the characteristics already mentioned in the Introduction, another factor that underscores the need for effective Web search strategy is the fact that most search engines index every word of a document. This method of indexing tends to greatly increase the number of results retrieved, while decreasing the relevance of those results, because of the increased likelihood of words being found in an inappropriate context. When selecting a search engine, one factor to consider is whether it allows the searcher to specify which part(s) of the document to search (eg. URL, title, first heading) or whether it simply defaults to search the entire document.
The most productive searches are those where the information seeker has spent time working out a search strategy before going online. The strategy is a pre-requisite for anyone attempting exhaustive searching, such as those embarking on a PhD, and recommended practice for any student wishing to conduct an efficient search and avoid frustration caused by low retrieval. In situations where connect time is charged for a search strategy is essential to prevent escalating costs.
Work out your specific information need and identify the different major concepts and alternatives. For example, the topic Inorganic fertilizers divides into two main concepts:
Put ideas on paper in natural language.
Examine each concept to find as many synonyms and terms as you can think of, and group the related items together to provide the basis of a structure for searching:
Inorganic fertilizers fertilization soil Soil fertilizers fertilized plants Fertilizers producing factories factories
Consider the levels - the amount of information you want, any limitations by date, language, etc and add these qualifications to the structure.
Boolean logic
Boolean logic is the term used to describe certain logical operations that are used to combine search terms in many databases. The basic Boolean operators are represented by the words AND, OR and NOT.
AND
If you need to pose a more specific query, use the boolean operator AND, which limits results to those items that contain both (or all) of the search terms in your query. Again using the two words from the example above, the search query would retrieve only those items containing both words in the same item
As the shaded area shows, this search query would return a much smaller set of hits, and the items would be more applicable to the field of inorganic fertilizers. To demonstrate the difference between the OR and the AND operator, we ran the two searches above using Internet. The search query Inorganic fertilizers OR Soil fertilization returned over 31,000 items, while the query Inorganic fertilizers AND Soil fertilization returned 176 items.
OR
The OR operator is useful for the first phases of a search, when you are not exactly sure what information is available on your topic or what words are used to categorize it. When used between two words, the OR operator instructs the search tool to retrieve any record containing either of the words. For instance, the search query would retrieve items containing either the word "fertilizers" or the term "fertilization":
Once you view the types of items containing either word, you might want to narrow your search by dropping one term and confining your search to the other. For instance, you might find that the records indexed under the term "fertilizers" are more relevant to your research question than those indexed under "fertilization". Or, as in the example below, you might find that the items related to the specific field of "soil fertilization" must contain both words, not simply either one. Because OR is the boolean operator that returns the most "hits" (items meeting the search criteria), search queries containing OR are very broad and sometimes return items that are not relevant.
NOT
The last of the three most common boolean operators is the word NOT. The NOT operator is used to eliminate records containing a particular word or combination of words from your search results. For instance, if you are performing a general search on soil fertilization, you might wish to exclude items dealing with the very specific discipline of "fertilizers production". To make this exclusion, you could construct your search query as:
This search would return all items containing the word "fertilizers" except for those that also contain the word "organic.
When you visit a search site, always read the instructions or help file before beginning your search. Each search engine has different parameters for using upper- and lower-case letters and combining boolean operators. Another good method for refining your search is to run a few searches experimentally to see what results are returned. By browsing through your results list, you can determine whether or not your strategy is returning relevant items. Then, you can construct a search strategy using the boolean operators OR, AND, and NOT to improve your results.