Search Tips

 

 

Quick Search Guide

---

 beyond Keyword searching

Beyond searching in natural language across the Internet, many search tools unique to the "hidden Internet" allow the user to create more complex search strings beyond just entering keywords. These word combinations allow the user to increase the number of relevant hits while decreasing the number of irrelevant ones. Sometimes the search engines combine keywords automatically, or make assumptions for the user, based on how the words are entered. Knowing what the assumptions are and how to manipulate, or override, them can make searching in a keyword environment more efficient.
Before using one of the Library's online databases for the first time, it is wise to look at the search hints that most of these purchased database vendors' search engines provide.

Listed below are some of the more common means of combining keywords for searching. These are general explanations only. It is up to the user to determine how a specific database search engine provides these capabilities.

BOOLEAN OPERATORS -- to expand or restrict your search results

 

PROXIMITY OPERATORS --being more precise for better results limits the search

Boolean operators can be very useful in finding documents with certain keywords in them. However, there are times when more precision is needed, as when the words are very common (i.e. "gun control"). Being able to tell the system that the words must be located near each other, perhaps together in the title or subject heading, or just near each other in the text, can be very useful. Proximity specifications make this precision possible. The examples below are general descriptions only, as each database handles proximity differently.

• infallible (W) rule --The words must be adjacent to each other and in that order
• deep (3W) structure -- The words must be within 3 words of each other and in that order
• relevance (N) judgment -- The words must be adjacent but can be in any order
• control (3N) guns -- The words must be within 3 words of each other, but can be in any order

To indicate proximity within a particular field, the field code would be linked to the search string. For example,
    "ti: control (2N) guns"
would search for a title with "guns" and "control" within 2 words of each other. This spatial relationship between terms is present in most databases. However, the exact characteristics of the features and the way they are executed in individual databases does vary, so be alert.

 

Truncation

--using only parts of words to expand a search

Truncation allows the user to search on a piece of a longer word or phrase, usually its left-most portion. It is like using a "wild card" in place of one or more letters in a word. The mark used to indicate truncation varies, but the concept is the same. If the question mark (?) is used to indicate truncation, then the search term

     libr ?

will retrieve the set of postings associated with records that contain any term beginning with the character string "libr," including

     library, librarian, libraries, librarianship, Libra, libration, and libreto.

Used wisely, truncation can save much keying time.  However, truncating too much can generate many false hits, or irrelevant documents. The term

     lib ?

for example, would retrieve records containing the library terms listed above, but would also retrieve

     liberty, liberal, libido, Libya, and libel, to name but a few.

Embedded truncation, or masking, is also possible with some systems. Masking allows a "wild card" symbol to be used in place of one letter in a word. It is useful when the spelling of the word is not certain, or when the words being searched differ by only one letter

     capit*l for either "capital" or "capitol"

     wom*n for either "woman" or "women"

Using truncation and masking generally expands the number of items retrieved, so it should be used carefully to prevent numerous irrelevant items.

Stop Words

--don't use these words in a search string!

Boolean Operators, proximity specifications, and truncation are tools a searcher can use to increase the relevance, and/or decrease the irrelevance, of search results.  Stop words, however, help the search engine work more efficiently by eliminating frequently used, generally valueless, words from being searched.  As with the tools listed above, each database search engine has defined its own list of stop words. Some of the most common stop words include

  a   for   on   an   from   the   and   in   to   by   of   with

The only time stop words become an issue in searching is when the user wants to include one of the words in a phrase. The technique for this inclusion depends on the database. In some databases, the phrase

     Gone with the wind

can be found by searching the phrase

     "gone with the wind" as a bound phrase (inside quotation marks).

Other databases would require the phrase to be searched using proximity specifications, i.e.

     gone (2W) wind

Before using very common terms in a phrase, it is important to find out how each search system deals with stop words.