Expand	

         The Expand option, located in the Search Window, is 
         designed to aid you in finding information which may 
         not be easily retrieved by standard searching.  The 
         Expand option assists in searching for issues over 
         which you do not have complete understanding.

         The Expand option displays a list of words that are 
         related to the word, words, or document identified for 
         further analysis in the Expand option.  The Expand 
         option is a statistically generated, extended thesaurus. 
 
         To activate the Expand Search capability, place the
         text (one or more words, paragraphs, documents, ID #s) 
         in the Search Window, then move your mouse pointer to 
         the Expand Button and click.  This action will 
         automatically open a window displaying the word 
         "Expanding."  After a short period of time, the window 
         will normally fill with an array of terms identified 
         as related in some way to the original text.  The words   
         are displayed in priority order from left to right 
         across a row of words and then down to the next row.  
         The number of words displayed by Expand will normally 
         correspond to the size of the Expand Window.  You may
         resize this window at any time to control the number of
         words to be displayed.

         The list of words produced by the Expand can be used in
         several ways.  These words can improve the recall of 
         topically related documents by suggesting alternative 
         terms that may be appropriate to use in retrieving 
         relevant documents.  Additionally, the Expand terms can   
         also help to identify subtopics of particular interest.  

         The Expand option has some limitations when used with 
         a small database such as those found on Windows Personal
         Librarian.  Since Expand is a statistical operation, its
         usefulness and accuracy depends, in part, on the amount of
         material in the database specifically related to the topic
         of interest.  With very small databases, there will be a
         significant occurrence of noise words, i.e., words that
         scored well but are not particularly related to the desired
         topic.  The larger the amount of material in a database on
         a given topic, the fewer the number of noise words.