LIBRARY HOME | FIND BOOKS | FIND ARTICLES | SEARCH BY SUBJECT | LIBRARY SERVICES | SEARCH THE WEB | HELP

Principles of Searching

Natural language searching vs. Structured Searching

Searching is a language game and it is key to remember that though your adversary (the database) may have a lot of information in it, it is essentially dumb. The computer sees words a series of letters rather than the concepts they represent. So, for instance, a search on the concept of cats, using the word cats, will not retrieve items where the words "feline" or "cat" occurred instead of cats. Unlike us, the computer does not have the ability to understand the relationships between these words. The way we ask questions is called natural language. Computers currently can not understand natural language, they must be questioned via "structured searching". Luckily we have a set of conventions built up for structured searching that are very easy to use. 

Boolean Logic

Boolean logic refers to the logical relationship among search terms:

• OR
• AND
• NOT

Each operator can be visually described by using Venn diagrams, as shown below.

OR Operator - Use OR to retrieve records or pages that contains EITHER of two or more terms. OR is frequently used to search for synonymous terms or a variety of specific ways of expressing a general concept

Cats OR Felines	Cat OR Cats OR Felines

I need information about cats.

Boolean logic:    OR

Search:    cat OR cats OR felines

AND Operator - The AND operator makes sure ALL the terms one requests appear in the search results.

Cats AND Fleas	Cats AND Fleas AND Side Effects

Query:    I want to learn about the side effects of fleas on cats.

Boolean logic:    OR, AND

Search:    (cats OR felines) AND behavior AND side effects

NOT Operator - Use NOT to ensure that certain words won't appear in the search selections

Query:    I'm interested in jaguars, but not the car.

Boolean logic:    NOT

Search:    jaguars NOT cars

Truncation

An easy and important way to broaden your search is to search for variant endings of a word. You truncate the root of the word and search for all the variant endings of that root. It is often indicated through a * or a ?.

Unlimited truncation	Single character Truncation	Internal truncation  or masking
comput$ finds computer computers computing computerize	Car* finds cars care but not caring	Wom!n finds women woman

Steps to Formulating and Conceptualizing a Search

1. Identify concepts
2. List keywords for each concept and their synonyms
3. Specify the logical relationships among the keywords.

Proximity, Adjacency, or Phrase Searching

Specifies that two or more terms must appear close to one another, e.g. "adjacent" to each other (in any order), "within" the same subject heading, sentence, or paragraph, or "near" (e.g. within 5 words of) each other. This is especially useful for searching full text databases. Specific commands for proximity searching vary widely between databases, so consult each database's help screens to use them properly.

Stop Words

In most databases words of little subject significance that occur very frequently are excluded from searching. They can, however, be included in phrases by putting them in quotes: e.g. "near" death experience?" Examples of stop words include of, and, the, he, her, etc.

Limiting you Search

Most of databases, but not the search engines, include limiting options to cut down your retrieval. Depending on the database you can limit by:

• year of publication
• gender
• type of document
• language
• geographic location
• age
• lots of other things

Subject Headings VS. Keyword Searching

Subject searching involves searching the subject headings (sometimes called descriptors) of records in a database. Most databases have subject headings  for each record that is indexed. These headings are in the subject or descriptor field. The database producer assigns subject headings to books and articles from a list of terms used specifically for that database. This list, called a database thesaurus, ensures that all items about the same topic have consistent subject headings. Users can then retrieve all the items on a topic using a single term, even when there may be several good ways to state a concept.

• Must, however, think like the index creator.
• Not every topic fall neatly into a precise subject heading that really reflects what is being covered.
• If you are not satisfied with results of subject searching, the next step is to try keywords.
• Keywords will usually give more hits, though they will sometimes include false hits because computer looks for words anywhere in the document and they may not have a relationship to each other or very high relevance

Keyword Searching

A keyword search retrieves words or phrases from several important fields of the records in a database. In most databases a keyword search finds words in fields that have descriptive content, such as title, subject/descriptor, and abstract. In some databases, additional fields may be included in the keyword search. A keyword search usually retrieves more items than a subject search, but they may not all be relevant. In a keyword search you can retrieve a number of irrelevant items because the computer is looking for the exact word you typed, not for the meaning or context of the word.

For example, a search on AIDS will retrieve items on

• aids for the hearing impaired
• school aids
• AIDS (the disease)

Keyword search is the best method when

• there is no subject heading for your topic
• the subject heading is too general or too specific
• you are searching for a new trend or concept
• the database does not have subject headings

When searching by keyword, use only significant words, not common words, such as the, of, an, and that. Such words may be stop words, words that occur too frequently to search.

Tips on Conducting Effective Searches

1. Read the directions for your database.
2. Know when to use the "advanced" mode. Do so in most cases.
3. Include synonyms or alternate spellings in the search statements and connect these terms with Boolean logic.
4. Check your own spelling.
5. If the results are not satisfactory, repeat the search using alternative terms.
6. If there are too many results, or results that are not relevant:
7. Add concept words
8. Use vocabulary that is more specific to the topic
9. Narrow the search to an individual field in a record or parts of the Web page such as Title, Summary, First Heading, etc.
10. Use the Boolean NOT to keep out records containing terms not wanted

For help anytime consult a Reference Librarian at the Reference desk.