[Solar-L] Use of the * in OPAC searches

[Hensler, Deborah]

A library had a question about why a search for a videorecording wasn't
working as it seemed it should when using the wildcard * . 

In doing searches for the DVD "Horatio Hornblower [videorecording] : the
new adventures," some of the searches we tried are: 

1. Search Terms: Horatio Hornblower Videorecording
2. Search Terms: Horatio Hornblower Video*
3. Search Terms: Horatio Hornblower Videor*
4. Search Terms: Horatio Horn*
5. Search Terms: Horatio Hornb*

The question I sent to the Help Desk is Why do searches  2, 3 and 4
return "no records found" as an AND query and then default to an OR
query, while search 5 jumps to the correct record?

The HD responded that "the single truncation "*" matches only up to 5
characters.  To get better results, use the double truncations since
this matches any number of characters (see description in Manual
#101243)."

So, the reason it doesn't find the video when using "Horatio Hornblower
video*" is that the single wildcard * will only match up to 5 additional
letters in a word. Because videorecording has 9 more letters than video,
it doesn't find it. The way around this is to use 2 wildcards, as in **,
and it will match any number of additional letters. 


The page referenced in the Manual includes this info:

?	Matches any single character in the specified position in the
word. For example, "anders?n" will match both "anderson" and "andersen".
The '?' wildcard may only appear after at least two characters.

 
*	Matches up to five non-space characters, starting at the
specified position in the word. For example, "inter*" will match
"internal" and "internet", but will not match "international". The '*'
wildcard may also be embedded in a search string. For example, "colo*r"
would match both "color" and "colour".

The '*' wildcard may only appear after at least two characters. For
example, "n*" would be rejected, but "ne*" would be accepted (but would
likely produce an unmanageably large result set).

 
**	Matches any number of non-space characters, starting at the
specified position in the word. For example, "inter**" will match all
words that begin with "inter" (e.g., "internal", "internet",
"international", etc.).

As is the case with the '*' wildcard, the "**" wildcard may only appear
after at least two characters (e.g., "p**" would be rejected).


 
In the Horatio Hornblower examples, a double asterisk after 2,3 and 4
would have matched the title. 

Hopes this help anyone out there who sometimes wonders why some searches
don't work when it seems they should. 

Deb
_______________________
Deborah Hensler
Library Systems Specialist
Maine State Library
64 State House Station
Augusta, ME 04333
207-287-5620 or 1-800-322-8899
207-592-1825 - mobile
207-287-5624 - fax