mining search logs for index terms

[Jeff.Kuntzman at UCHSC.edu]

Although our efforts have not been very scientific,
we regularly scan our search logs and have added several
web pages based on that.
One example:
"Renew books" is an instance of a constantly
recurring search term - we now have a web page on our library 
web site about renewing books. (Like many libraries the actual 
process of book renewal takes place on the webpac, not the web 
site.)
Jeff Kuntzman
Denison Memorial Library
University of Colorado Health Sciences Center

>Have you considered mapping the terms you uncover in your 
>search logs directly into your thesaurus?  I've been doing 
>that for a database of the most commonly searched for websites 
>on our intranet and its been very successful.  We use the thesaurus 
>tree for browsing (broad, generic terms), and add in see references 
>for searching (terms taken from the search logs). Is anyone else 
>mining their search logs for index terms? 
> Lydia Bauer Bristol-Myers Squibb Company Knowledge 
>Integration Resources lydia.bauer at bms.com