[Corpora-List] corpora at hd.uib.no

Amruta D. Purandare amruta at cs.pitt.edu
Fri Dec 24 02:20:31 UTC 2004


Hi,

You may want to try SenseClusters -
http://senseclusters.sourceforge.net

Its basically a text clustering package but does have some code
for LSI/LSA.

Regards,
Amruta

___________________________________________________________________
Amruta Purandare		    amruta at cs.pitt.edu
Intelligent Systems Program	    http://www.cs.pitt.edu/~amruta
University of Pittsburgh	    (412)-657-1318
___________________________________________________________________

On Thu, 23 Dec 2004, [ISO-8859-1] Leif Grönqvist wrote:

> Hi all,
>
> I have a problem with my thesis work: the LSI-software I have tried
> out is not perfectly stable.
>
> - Infomap from Stanford is nice but does not scale up since they don't
>   use a compact format for sparse matrices. There are also some bugs
>
> - The free software from Telcordia handles huge corpora but has a lot of
>   bugs...
>
> So, does anyone have an idea of a stable package? I want the document
> handling parts too, not just the SVD.
>
> Merry Christmas!
>
> Leif Grönqvist, GSLT, leifg at ling.gu.se, www.ling.gu.se/~leifg, 031-821515(home)
> School of Mathematics and Systems Engineering, Växjö University 0707164380(mob)
> Department of Linguistics,  Göteborg University, +46 31 773 1177, 773 4853(fax)



More information about the Corpora mailing list