You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@nutch.apache.org by Erik Hatcher <er...@ehatchersolutions.com> on 2005/12/01 01:12:17 UTC
Re: "Good man" is Different than "Man good" in Nutch?
On 29 Nov 2005, at 22:41, Victor Lee wrote:
> ok, now I remembered something from the book Lucene in Action, it
> said something about "word distance". So that's why they returns
> different results. But still, when I remembered when I went to
> Google Adwords and get the new Maximum CPC estimates for phases
> containing same words but with different orders, they always
> treated them to be the same by showing the same statistics.
> Why?
What does the "explain" link tell you about the differences? If I
recall correctly, Nutch generates a PhraseQuery for your terms and
OR's that in with TermQuery's, which would favor documents that have
the terms in the order listed. The explanation will tell all.
Erik
>
> Victor Lee <vi...@yahoo.com> wrote: Hi,
> When I went to mozdex.com which is using Nutch, I realized that
> the search term "good man"(no double quotes in actual search term)
> returns different search result than the search term "man
> good" (also no double quotes in actual search term). I went to
> Google and they are doing similar thing. Why? I thought that all
> terms are connected with AND by default, so they should return the
> same search result.
>
> Many thanks.
>
>
>
> ---------------------------------
> Yahoo! Music Unlimited - Access over 1 million songs. Try it free.
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com