14 Oct
2011
14 Oct
'11
11:44 a.m.
Thanx Doru.
I am looking now at Visual works and porting missing methods.
regards
Usman
On Fri, Oct 14, 2011 at 8:49 AM, Tudor Girba <tudor(a)tudorgirba.com> wrote:
Hi,
The code is in the Moose-Algos-FormalConceptAnalysis package from the Moose
image.
The algorithm is ported, but there are missing extensions for the Integer
hierarchy related to bit logic.
If someone wants to port the rest, just run the below command, and then
search for the methods in the VW version. They should be published as
extensions in Moose-Algos-FormalConceptAnalysis.
MalFormalContext mammals2 computeConcepts
Cheers,
Doru
On 13 Oct 2011, at 19:53, Stéphane Ducasse wrote:
--
www.tudorgirba.com
"Not knowing how to do something is not an argument for how it cannot be
done."
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev
doru
do you know if the library of adrian for FCA has been ported to Pharo?
Stef
Begin forwarded message:
> From: Usman Bhatti <usman.bhatti(a)gmail.com>
> Date: October 13, 2011 5:28:18 PM GMT+02:00
> To: gabriela.b.arevalo(a)gmail.com
> Cc: Stéphane Ducasse <stephane.ducasse(a)inria.fr>fr>, Nicolas Anquetil <
nicolas.anquetil(a)inria.fr>
> Subject: looking for a FCA library in
Smalltalk
>
> Hello Gabriella,
>
> I am working with Stephane and Marianne on a paper related to FCA
patterns.
The idea is to identify a few patterns in FCA lattices that would
reduce lattice exploring task to searching and looking at a few of these
patterns. We describe the combination of nodes and subgraphs in the lattices
as patterns. When identifying patterns, we took care of the following
things:
>
> 1. the patterns should be generic (i.e. not related to a particular
context)
> 2. the patterns should produce little
false-positives (that is the nodes
that are not important in lattices should not
appear in candidates for
evaluation)..
>
> we presented our paper in wcre 2011 but it got rejected. Now, we are
reviewing
the paper to present it in CSMR 2012.
>
> The purpose of the mail is that we are looking for a tool (specifically
in
Smalltalk) that would automate the process of patterns identification in
FCA. So, I am looking for a library in FCA that:
>
> - extracts program info from smalltalk (if possible)
> - calculates lattices based on program info
> - permits to do queries on lattices (top, bottom, a concept with
objects,
attributes, etc.)
> - find horizontal decomposition and modules
in lattices.
>
> Would you have a knowledge of such a library?
>
> If you wish I can send you the paper to have a look at it. There is a
tool in
Java called Colibri that allows to do some of these queries on
lattices:
http://code.google.com/p/colibri-java/
but its written in Java and I am looking for something in Smalltalk.
thanx in advance,
Usman
_______________________________________________
Moose-dev mailing list
Moose-dev(a)iam.unibe.ch
https://www.iam.unibe.ch/mailman/listinfo/moose-dev