Most databases by default search for keywords found anywhere in the document. If you were to enter the term:
without any qualifiers or delimiters, the search results would include
Stevens, Timothy. (2012). Christopher Columbus : a biography. London: Notting Hill Press
as well as:
Hill, Christopher (1985). Arrest of Ships. London : Lloyd’s of London Press Ltd. , 1985
And if you were searching a database of full-text documents, depending on search engine defaults, your search terms might pull up a document that has the word “Christopher” on page 4, and the word “Hill” on page 444. Your results would be “comprehensive” in catching all documents with your terms used anywhere, but they would contain a great deal of materials that are not “relevant” to what you are looking for. In the library Discovery Service, a search this indiscriminant could retrieve tens or even hundreds of thousands of documents — far too much for you to wade through in search of items useful to your research.
When not otherwise specified by the user, a search term with multiple words will by default AND those words together. Any item in the results must include “Christopher” AND “Hill.” The boolean operator AND will always be a narrowing influence because ALL words in the term must appear in the document. Of course, when you are searching full-text databases and/or documents with millions of items, a simple AND search might not seem to be limiting your results, but if you were to search for “Christopher” without adding the word “Hill”, or vice-versa, your search results would almost certainly be greater than when using both terms.
OR, on the other hand, will increase the “comprehensiveness” of the results over using any single term. The document either contains the word “Christopher” OR it contains the word “Hill” (or both). If you find that your search is too narrow, you can expand it by using OR. For example, a search for:
Author=”Christopher Hill” AND Title=(“Law” OR “Ships”)
would pull up Christopher’s book on Maritime Law and his book on Arrest of Ships. Note the parenthesis around the OR phrase. When the advanced search of a database allows you to specify different fields in which to place your terms, you would likely not need parentheses. If, however, you were searching in a single box, your results could be very different without parentheses.
Author=”Christopher Hill” AND Title=”Law” or “Ships”
Would retrieve anything by Christopher Hill with the word Law in the title, but it might also retrieve anything with the word Ships in the title, whether or not it was authored by Christopher Hill. The parantheses ensures the logical groupings.
Not might not alter your search results, but when it does it will limit them. If for example, you searched:
Author=Christopher Hill NOT Publisher=World Maritime University
The results would be the same, because there are no works authored by Christopher Hill and Published by WMU. On the other hand:
Subject = Piracy NOT Copyright
Would likely limit your search results in an appreciable and relevant way.
One technique for narrowing your search is to limit your terms to a single field. Many databases have an option to limit your terms to Author, Title or Subject fields, meaning they would not be searched across the document, but only the fields you specify. Often a database will have an Advanced search option that lets you put different terms in different fields so that you can limit your search, for instance to Author=Christopher Hill and Title=Arrest of ships.
Phrase searching refers to searching for words in exact order as a single string, rather than searching for each word separately. In most databases, a phrase is designated by quotation marks as in:
“United Nations Convention on the Law of the Sea”
This exact phrase must appear in the document in order for it to be returned in our search results. Phrase searching is an important tactic when you need to limit your results.
Truncation and wildcard
Many databases allow you to provide a character to represent one or more characters following your term, in order to catch variations of the word and plurals. In most databases (but not all), it is represented by an askerisk symbol (*). For instance:
Would retrieve documents containing the word sefarer, seafarers, seafaring.
A wildcard is a substitution of a single character (or no character) and can be used to catch variations in spelling.
Would retrieve documents with the word “Labor” or “Labour.” Wildcard symbols might not be avialable in some databases, and the choice of characters can vary according to database, so be sure to look this up in the help. Truncation and wild cards are fairly surgical means of expanding your search results.
Many of the databases the library offers have hundreds of thousands of records and it is as much an art as it is a science to try to sift through them to limit your results only to pertinant items without missing anything of relevance. It is usually tradeoff between precision and comprehensiveness. This is where putting the limiting and including techniquest together in combination can be very useful, partularly with the discovery service, where the possible items retrieved can be in the tens of millions. Although the example below shows what is going on “under the hood,” (and can still be done by experienced users in databases offering “command line” searching), fortunately, most web-based interfaces will offer appropriate text fields and pull down menus to achieve the same results..
|( ( SU(“PIRACY”) AND SU(“MARITIME INDUSTRY”) OR SU(“SEA TRANSPORT”)) )
|Find documents with the subject heading of “piracy” that also have either the subject heading “maritime industry” or the subject heading “sea transport”
|OR SU(“SHIP HIJACKING”)
|or has the subject heading “ship hijacking.”
|AND (YR(>2009)AND tanker*
|The document must also be published after the year 2009 and must have the word tanker or tankers anywhere in the text of the document or metadata.