Mailing-List: contact dev-help@mahout.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@mahout.apache.org
Received-SPF: neutral (athena.apache.org: local policy)
From: Isabel Drost <isabel@apache.org>
Reply-To: isabel@apache.org
To: dev@mahout.apache.org
Subject: Re: GSOC 2012
Date: Mon, 19 Mar 2012 22:10:55 +0100
User-Agent: KMail/1.13.7 (Linux/3.0.0-1-amd64; KDE/4.6.5; x86_64; ; )
References: 
 <CAOEqKat08d-956yi+9zufiWv06JXoYNxdwrq4V-Fa4XZQM-hmQ@mail.gmail.com>
 <CAPQyphv1_iW=YAQJzaStgP_KObBQgnSV6xF-mJ9NcRWRMAg5+g@mail.gmail.com>
 <CAOEqKauJoHEgk4qnShHCdDF8sKCThxdZO+JKtJAAZ0ubhY0FTA@mail.gmail.com>
In-Reply-To: 
 <CAOEqKauJoHEgk4qnShHCdDF8sKCThxdZO+JKtJAAZ0ubhY0FTA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/signed;
  boundary="nextPart5721415.ofleo1skSQ";
  protocol="application/pgp-signature";
  micalg=pgp-sha1
Content-Transfer-Encoding: 7bit
Message-Id: <201203192210.56310.isabel@apache.org>

--nextPart5721415.ofleo1skSQ
Content-Type: Text/Plain;
  charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

On 19.03.2012 Dirk Weissenborn wrote:
> What could be possible projects for this year?

General advise would be to look for open JIRA issues. There are a few known=
=20
issues with existing implementations. Also for some algorithms integration =
and=20
API design could be improved. In the end the general advise on how to contr=
ibute=20
and become a committer applies: Find a topic that you are yourself interest=
ed in=20
and work on that.


> Or should the student
> provide a proposal for an improvement?

=46rom previous experience students who came up with their own proposals ha=
ve been=20
most successful - both in applying but also in succeeding. Keep in mind to =
limit=20
the scope of your project to a manageable size: You can only design, test,=
=20
implement, document and provide examples for so much code.


> I would probably be interested in
> working on a project for this years GSOC. I am also currently in contact
> with dbpedia spotlight developers for GSOC. They are interested in a
> topical classifier for a given plaintext. I saw, that something similar w=
as
> already implemented in mahout.

I think working on the intersection of two projects could be valuable for b=
oth=20
sides. Sounds like an interesting idea to me.

Isabel

--nextPart5721415.ofleo1skSQ
Content-Type: application/pgp-signature; name=signature.asc 
Content-Description: This is a digitally signed message part.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iEYEABECAAYFAk9noOAACgkQPJpCBAwIhbSi0wCgg1B63zc3YlDRMDmrcoqXAk1g
XZQAoKtOC0pQcoshfpEnKTuku/wUmpur
=pcqx
-----END PGP SIGNATURE-----

--nextPart5721415.ofleo1skSQ--