[KinoSearch] get doc/query similarity

[jack_tanner at yahoo.com]

Ping? Still trying to compute similarity of two indexed docs... A weighted cosine or some such.

Thanks in advance.

----- Original Message ----
> From: "jack_tanner at yahoo.com" <jack_tanner at yahoo.com>
> To: KinoSearch discussion forum <kinosearch at rectangular.com>
> Sent: Friday, April 11, 2008 3:07:05 PM
> Subject: Re: [KinoSearch] get doc/query similarity
> 
> From: Marvin Humphrey 
> >
> > Let's assume you mean a term, for the sake of getting things started.   
> > Let's also assume that you don't really mean "one specific document",  
> > even though that's exactly what you said. :)
> 
> Thanks for that example. Let me be more clear about what is desired: I need to 
> compute the similarity of two indexed documents. I'd like it if the metric was 
> more sophisticated than mere term overlap. At a minimum, it could be Jaccard 
> (i.e., doc length-normalized term overlap). It would be preferable to have 
> something that takes corpus statistics into account. For example, if in my 
> corpus some term T has high TF and low IDF (occurs often and in many docs), then 
> such a term could be downweighted. Could you suggest a way of doing this? 
> Ideally with KS 0.162?
> 
> > Interesting.  I received your private email and wrote back.  Maybe  
> > hotmail is blocking rectangular.com or something.  AOL blockaded me  
> > once because the previous tenants on the Comcast IP block  
> > rectangular.com got assigned to weren't good netizens.
> 
> I never got your response at my hotmail address, not even in the spam folder. If 
> you like, I could forward a complaint to Hotmail's postmaster. Please send a 
> test e-mail to my @yahoo and cc my @hotmail, and I'll forward that along.
> 
> 
> 
> 
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around 
> http://mail.yahoo.com 
> 
> 
> _______________________________________________
> KinoSearch mailing list
> KinoSearch at rectangular.com
> http://www.rectangular.com/mailman/listinfo/kinosearch
> 




      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ


_______________________________________________
KinoSearch mailing list
KinoSearch at rectangular.com
http://www.rectangular.com/mailman/listinfo/kinosearch