Forum Moderators: open

Message Too Old, No Replies

Google says common word - not included in search

search results different with or without word

         

LogicMan

9:17 pm on Aug 4, 2003 (gmt 0)

10+ Year Member



I do not care about the common word(s), I want to understand the factors that make a differnce.

Google says "a" is a very common word and was not included in your search BUT if I leave it out of my search the results are different. I also tried with "in", "the", etc.
For example, widgets in the factory vs. widgets factory

If Google is really not including it in the search, what part of their formula can be making the results different.

jcoronella

10:10 pm on Aug 4, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Also try checking 'buy a widget' and then 'buy +a widget' and then again 'buy widget'

The second one forces google to include the 'a', and all are different.

There is a certain part of the algorithm dealing with proximity of multiple words, and that is probably negated or lessoned with 'buy a widget'. In a similar way that the two queries 'blue widget' and 'widget blue' differ.

takagi

3:37 pm on Aug 5, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



The common word is replaced by an asterisk ('*') which acts like a wildcard. Although google still writes

Google does not use "stemming" or support "wildcard" searches
Google Basics [google.com]

they do support both to some extend. A kind of stemming was introduced yesterday; see the thread: New Google search operator [webmasterworld.com].

The asterisk as a wildcard has been active for some time now. It can only be used to replace a word, unfortunately not to replace a part of a word. It makes most sense to use it in a qouted string. Using an asterisk in a query without qoutes can change the order of the SERP without changing the number of results. Just as for changing the location of the keywords of the query string.

The asterisk stands for one word, so a search for

"new * widgets"

will find 'new blue widgets' and 'new handmade widgets' but not 'new widgets' or 'new blue painted widgets'. You can also use 2 asterisks to replace 2 words. So the search for

"new * * widgets" 

will find the 'new blue painted widgets' but not 'new blue widgets'. To find all the posts made by GoogleGuy last month, you can search for

site:www.webmasterworld.com GoogleGuy "July * 2003 utc"

However sometimes his name is also used in a message written by another member, so searching on the day he joined is better:

site:www.webmasterworld.com "joined-Oct 8, 2001" "July * 2003 utc"

Fiver

6:06 pm on Aug 5, 2003 (gmt 0)

10+ Year Member



basically you should be able to search for

"keyword1 stopword keyword2"

and get the same results no matter what the stopword [webmasterworld.com]

there's other talk on this, I asked this question here a while ago [webmasterworld.com]

RBuzz

8:43 pm on Aug 5, 2003 (gmt 0)

10+ Year Member



Google says "a" is a very common word and was not included in your search BUT if I leave it out of my search the results are different. I also tried with "in", "the", etc.
For example, widgets in the factory vs. widgets factory

If Google is really not including it in the search, what part of their formula can be making the results different.

--

Google seems to weigh the order of the query words in how it provides results. In other words, if you search for find a widget, it's going to look for "Find a widget" first and provide those results first -- even if a is a stopword. (At least that how it seemed to me after a LOT of experimenting. I am not privy to Google's secrets.)

There are tools out there that will do proximity searches on Google even though Google itself doesn't offer that feature -- Google for
"Google API Proximity Search."

> The common word is replaced by an asterisk
> ('*') which acts like a wildcard. Although
> google still writes Google does not use
> "stemming" or support "wildcard" searches"

It's the difference between full-word wildcards (which is what Google offers) and stemming (which AltaVista for example offers, but Google does not.)

RBuzz