Hi,
The similarity measures other than cosine - match, jacc, dice etc. do
not use the frequency information for their computation, they just need
binary information about features presence/absence which is supported by
bitsimat.pl.
Thus if the requirement is to cluster in similarity space using
similarity measure other than cosine then you only need binary context
representation and then proceed with using bitsimat.pl.
Regards,
Anagha
Hi Ted,
Is there planned support for different measures of similarity in
senseclusters' simat.pl?
Right now there's the cosine function, but I'd be curious to compare
results based on jaccard and other measures as well.
I noticed that you have some choices when using CLUTO, but i'm working
most of the time on raw similarities.
If there's evidence or experiences that suggests one measure over
another, I'd also be interested finding out more about that
kind regards,
s.
-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
senseclusters-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/senseclusters-users
