AltaVista bafflement - "santa fe" "Santa Fe"

[Laura Cohen]

I find the Santa Fe results on AltaVista especially interesting in the
light of AltaVista's automatic phrase detection. Here is a quote from the
November 4, 1998, Search Engine Report
(http://www.searchenginewatch.com/sereport/9811-altavista.html) : 

           "We have built a huge dictionary of several million
           phrases," said Monier ( AltaVista's chief technical officer).
           "Essentially, the need for quotes has gone down to zero. 
            If you look for New York, you don't need to quote New York."

The search that Miriam Bobkoff related to this list today isn't the only
inconsistency you'll find at AltaVista. For example, AltaVista is supposed
to default to OR logic, except in cases in which you search on a phrase
that appears in its phrase dictionary. Right? Well...

Try this search:

	title:Mars    host:nasa.gov

This search defaults to AND logic, as all results come from the nasa.gov
site. If the syntax were applied in a consistent manner, I should be
required to place a plus sign (+) in front of each search statement:

	+title:Mars    +host:nasa.gov

While it may make excellent sense to apply AND logic to my original search,
this does violate AltaVista's stated search rules.

When teaching about search engines, I always tell my audiences to expect
inconsistencies. I've even found the tried and true Infoseek to act up on
occasion.

-- Laura Cohen

>Man, this is REALLY wacky! I agree with Miriam; this is completely counter
>to what I expected of AltaVista, given my understanding of how they handle
>case-sensitive queries. I'm cc'ing this to the AltaVista support people.
>
>I tried the same case queries with "Santa Ana" and "santa ana" which 
>gave me more normal results ( 47513 and  68648 respectively.) I also 
>tried "san francisco" and "San Francisco", which also gave me expected
>results. Shaping the query as "santa fe" vs. santa;fe didn't make any 
>difference.
>
>Miriam, I did notice that if I went to the AltaVista advanced query screen
>( http://www.altavista.com/cgi-bin/query?pg=aqwhat=web&kl=en ) and 
>typed "santa fe" in the boolean query box, I got 303166 results. Typing
>"Santa Fe" got 154768 results. That may be a good workaround, though
>it doesn't explain why AltaVista is behaving this way.
>
>Puzzled,
>
>Tara

__________________________
Laura B. Cohen
Network Services Librarian
University at Albany  Libraries     
Albany, NY 12222                              518-442-3492 
LCOHEN at cnsvax.albany.edu       518-442-3567 (FAX)