Discussion:
Contributing in a New Topic : Recommender Systems
(too old to reply)
MIT SHAH
2014-02-02 14:39:58 UTC
Permalink
Hi,

I want to know whether there are algorithms on "Recommender Systems"
in scikit-learn. I didn't found this topic in documentation. If not, I
would like to contribute on this topic.
Please guide me.

Thanks !!
Andy
2014-02-02 17:18:16 UTC
Permalink
Hi Mit.

Some of the algorithms in scikit-learn could be use for recommender
systems, but there is no estimator for that.
We are currently hesistant in increasing the scope of scikit-learn, as
the core contributors are busy doing maintenance,
and we would rather focus on a stable 1.0 release.
If you'd like to see recommender systems in sklearn-style, you are more
then welcome to create your own gist / project
that is sklearn compatible and focuses on recommender systems. Once
other issues in sklearn are worked out and if we feel that
it fits, maybe it gets merged later on. Even if not, we can link to it
and it will be just as beneficial to the users as if it was in the core
project.

This is my opinion, other contributors might disagree, though ;)

Cheers,
Andy
Post by MIT SHAH
Hi,
I want to know whether there are algorithms on "Recommender
Systems" in scikit-learn. I didn't found this topic in documentation.
If not, I would like to contribute on this topic.
Please guide me.
Thanks !!
------------------------------------------------------------------------------
WatchGuard Dimension instantly turns raw network data into actionable
security intelligence. It gives you real-time visual feedback on key
security issues and trends. Skip the complicated setup - simply import
a virtual appliance and go from zero to informed in seconds.
http://pubads.g.doubleclick.net/gampad/clk?id=123612991&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
n***@masonlive.gmu.edu
2014-02-03 01:23:23 UTC
Permalink
This is in response to the thread on recommender system implementation in scikit-learn. I would also like to know if any of the scikit-learn contributors are willing to mentor a project which implements basic recommender system algorithms - collaborative filtering (user-based/item-based/matrix factorization) for Google Summer of Code. I feel the lack of a scalable, extensible open-source recommendation engine in python is an interesting gap to fill and would like to try my hand at it during GSOC. There are a couple of interesting problems to address in this case like how to design a recommendation engine that conforms to the design of scikit-learn package as much as possible. Some of the other challenges are implementing support for Sparse matrix operations.

Thanks,
Nikhil

________________________________
From: MIT SHAH [***@gmail.com]
Sent: Sunday, February 02, 2014 9:39 AM
To: scikit-learn-***@lists.sourceforge.net
Subject: [Scikit-learn-general] Contributing in a New Topic : Recommender Systems

Hi,

I want to know whether there are algorithms on "Recommender Systems" in scikit-learn. I didn't found this topic in documentation. If not, I would like to contribute on this topic.
Please guide me.

Thanks !!
Nick Pentreath
2014-02-03 07:19:00 UTC
Permalink
Tadej Štajner
2014-02-03 07:53:31 UTC
Permalink
<html>
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">There's also Crab, which doesn't really
conform to the sklearn API, but is still numpy/SciPy based.  <br>
<a class="moz-txt-link-freetext" href="https://github.com/muricoca/crab/">https://github.com/muricoca/crab/</a><br>
<br>
It doesn't seem to be actively maintained, but looks like it's
well enginneered.<br>
<br>
-- Tadej<br>
<br>
<br>
On 03. 02. 2014 08:19, Nick Pentreath wrote:<br>
</div>
<blockquote
cite="mid:CALD+6GM9DCw6Mrw=vpWb2tLqzXbXopjb6A2BnDgSUx-rV=***@mail.gmail.com"
type="cite">
<pre wrap="">There have been many people asking about contributing recommender systems
to scikit-learn, and generally the response has been that it doesn't quite
fit in with the library. Though it can be shoehorned somewhat perhaps, I
recommend you take a look at <a class="moz-txt-link-freetext" href="https://github.com/mendeley/mrec">https://github.com/mendeley/mrec</a>, which
implements a number of recommender algorithms, depends in part on
scikit-learn, and tries where possible to conform to the scikit-learn API.

Nick


On Mon, Feb 3, 2014 at 3:23 AM, <a class="moz-txt-link-abbreviated" href="mailto:***@masonlive.gmu.edu">***@masonlive.gmu.edu</a> &lt; <a class="moz-txt-link-abbreviated" href="mailto:***@masonlive.gmu.edu">***@masonlive.gmu.edu</a>&gt; wrote:

</pre>
<blockquote type="cite">
<pre wrap=""> This is in response to the thread on recommender system implementation
in scikit-learn. I would also like to know if any of the scikit-learn
contributors are willing to mentor a project which implements basic
recommender system algorithms - collaborative filtering
(user-based/item-based/matrix factorization) for Google Summer of Code. I
feel the lack of a scalable, extensible open-source recommendation engine
in python is an interesting gap to fill and would like to try my hand at it
during GSOC. There are a couple of interesting problems to address in this
case like how to design a recommendation engine that conforms to the design
of scikit-learn package as much as possible. Some of the other challenges
are implementing support for Sparse matrix operations.

Thanks,
Nikhil

------------------------------
*From:* MIT SHAH [<a class="moz-txt-link-abbreviated" href="mailto:***@gmail.com">***@gmail.com</a>]
*Sent:* Sunday, February 02, 2014 9:39 AM
*To:* <a class="moz-txt-link-abbreviated" href="mailto:scikit-learn-***@lists.sourceforge.net">scikit-learn-***@lists.sourceforge.net</a>
*Subject:* [Scikit-learn-general] Contributing in a New Topic :
Recommender Systems

Hi,

I want to know whether there are algorithms on "Recommender
Systems" in scikit-learn. I didn't found this topic in documentation. If
not, I would like to contribute on this topic.
Please guide me.

Thanks !!


------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.

<a class="moz-txt-link-freetext" href="http://pubads.g.doubleclick.net/gampad/clk?id=121051231&amp;iu=/4140/ostg.clktrk">http://pubads.g.doubleclick.net/gampad/clk?id=121051231&amp;iu=/4140/ostg.clktrk</a>
_______________________________________________
Scikit-learn-general mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Scikit-learn-***@lists.sourceforge.net">Scikit-learn-***@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/scikit-learn-general">https://lists.sourceforge.net/lists/listinfo/scikit-learn-general</a>


</pre>
</blockquote>
<pre wrap="">
</pre>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
<a class="moz-txt-link-freetext" href="http://pubads.g.doubleclick.net/gampad/clk?id=121051231&amp;iu=/4140/ostg.clktrk">http://pubads.g.doubleclick.net/gampad/clk?id=121051231&amp;iu=/4140/ostg.clktrk</a></pre>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">_______________________________________________
Scikit-learn-general mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Scikit-learn-***@lists.sourceforge.net">Scikit-learn-***@lists.sourceforge.net</a>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/scikit-learn-general">https://lists.sourceforge.net/lists/listinfo/scikit-learn-general</a>
</pre>
</blockquote>
<br>
</body>
</html>
NALINI RANGARAJU
2014-02-03 15:05:21 UTC
Permalink
About scikit-crab, I think the authors are re-engineering it and it is currently not open to the community for contribution. This is what someone said recently on the google group for crab. As far as I could tell (and I could very well be wrong), crab recommender system does not have support for sparse matrices which is an inherent property of recommender systems user/item rating matrices.

Sent from my iPhone
There's also Crab, which doesn't really conform to the sklearn API, but is still numpy/SciPy based.
https://github.com/muricoca/crab/
It doesn't seem to be actively maintained, but looks like it's well enginneered.
-- Tadej
Post by
There have been many people asking about contributing recommender systems
to scikit-learn, and generally the response has been that it doesn't quite
fit in with the library. Though it can be shoehorned somewhat perhaps, I
recommend you take a look at https://github.com/mendeley/mrec, which
implements a number of recommender algorithms, depends in part on
scikit-learn, and tries where possible to conform to the scikit-learn API.
Nick
Post by n***@masonlive.gmu.edu
This is in response to the thread on recommender system implementation
in scikit-learn. I would also like to know if any of the scikit-learn
contributors are willing to mentor a project which implements basic
recommender system algorithms - collaborative filtering
(user-based/item-based/matrix factorization) for Google Summer of Code. I
feel the lack of a scalable, extensible open-source recommendation engine
in python is an interesting gap to fill and would like to try my hand at it
during GSOC. There are a couple of interesting problems to address in this
case like how to design a recommendation engine that conforms to the design
of scikit-learn package as much as possible. Some of the other challenges
are implementing support for Sparse matrix operations.
Thanks,
Nikhil
------------------------------
*Sent:* Sunday, February 02, 2014 9:39 AM
Recommender Systems
Hi,
I want to know whether there are algorithms on "Recommender
Systems" in scikit-learn. I didn't found this topic in documentation. If
not, I would like to contribute on this topic.
Please guide me.
Thanks !!
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
Gael Varoquaux
2014-02-03 10:28:44 UTC
Permalink
Hi,

The discussion has indeed surfaced more than once:
https://www.mail-archive.com/scikit-learn-***@lists.sourceforge.net/msg08798.html
http://osdir.com/ml/python-scikit-learn/2012-06/msg00074.html
http://sourceforge.net/mailarchive/message.php?msg_id=30725712

And more recently:
http://sourceforge.net/mailarchive/forum.php?thread_name=CAFQAd-nyWHMSMvnYQUGUM03TK7hBpwVt%3DvBt%2BPH0-3wVtXjTpA%40mail.gmail.com&forum_name=scikit-learn-general

I think that the consensus that emerged from this discussion was that
recommender systems per say require a lot of domain knowledge and
combining machine learning primitives together. Thus recommender systems
per say seem out of scope of the project.

However, I think that we agree that basic primitives such as matrix
completion are in the scope.

So, in principle yes for a GSOC on the topic. But you need to find core
developpers motivated to mentor you, and they are all very busy. Also,
keep in mind that there is a lot of competition for the GSOC with
scikit-learn.

By the way: there has been an email thread on a GSOC on recommender
systems very recently. You need to read it.

Cheers,

Gaël
Post by
There have been many people asking about contributing recommender systems to
scikit-learn, and generally the response has been that it doesn't quite fit in
with the library. Though it can be shoehorned somewhat perhaps, I recommend you
take a look at  https://github.com/mendeley/mrec, which implements a number of
recommender algorithms, depends in part on scikit-learn, and tries where
possible to conform to the scikit-learn API.
Nick
This is in response to the thread on recommender system implementation in
scikit-learn. I would also like to know if any of the scikit-learn
contributors are willing to mentor a project which implements basic
recommender system algorithms - collaborative filtering (user-based/
item-based/matrix factorization) for Google Summer of Code. I feel the lack
of a scalable, extensible open-source recommendation engine in python is an
interesting gap to fill and would like to try my hand at it during GSOC.
There are a couple of interesting problems to address in this case like how
to design a recommendation engine that conforms to the design of
scikit-learn package as much as possible. Some of the other challenges are
implementing support for Sparse matrix operations. 
Thanks,
Nikhil 
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Sent: Sunday, February 02, 2014 9:39 AM
Subject: [Scikit-learn-general] Contributing in a New Topic : Recommender Systems
Hi,
     I want to know whether there are algorithms on "Recommender Systems"
in scikit-learn. I didn't found this topic in documentation. If not, I
would like to contribute on this topic.
     Please guide me.
Thanks !!
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/
ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
--
Gael Varoquaux
Researcher, INRIA Parietal
Laboratoire de Neuro-Imagerie Assistee par Ordinateur
NeuroSpin/CEA Saclay , Bat 145, 91191 Gif-sur-Yvette France
Phone: ++ 33-1-69-08-79-68
http://gael-varoquaux.info http://twitter.com/GaelVaroquaux
Olivier Grisel
2014-02-03 10:51:21 UTC
Permalink
Post by Gael Varoquaux
I think that the consensus that emerged from this discussion was that
recommender systems per say require a lot of domain knowledge and
combining machine learning primitives together. Thus recommender systems
per say seem out of scope of the project.
However, I think that we agree that basic primitives such as matrix
completion are in the scope.
+1 on my side.
--
Olivier
http://twitter.com/ogrisel - http://github.com/ogrisel
1970-01-01 00:00:00 UTC
Permalink
--001a1136bd706db3a204f17b5633
Content-Type: text/plain; charset=ISO-8859-1

There have been many people asking about contributing recommender systems
to scikit-learn, and generally the response has been that it doesn't quite
fit in with the library. Though it can be shoehorned somewhat perhaps, I
recommend you take a look at https://github.com/mendeley/mrec, which
implements a number of recommender algorithms, depends in part on
scikit-learn, and tries where possible to conform to the scikit-learn API.

Nick
Post by n***@masonlive.gmu.edu
This is in response to the thread on recommender system implementation
in scikit-learn. I would also like to know if any of the scikit-learn
contributors are willing to mentor a project which implements basic
recommender system algorithms - collaborative filtering
(user-based/item-based/matrix factorization) for Google Summer of Code. I
feel the lack of a scalable, extensible open-source recommendation engine
in python is an interesting gap to fill and would like to try my hand at it
during GSOC. There are a couple of interesting problems to address in this
case like how to design a recommendation engine that conforms to the design
of scikit-learn package as much as possible. Some of the other challenges
are implementing support for Sparse matrix operations.
Thanks,
Nikhil
------------------------------
*Sent:* Sunday, February 02, 2014 9:39 AM
Recommender Systems
Hi,
I want to know whether there are algorithms on "Recommender
Systems" in scikit-learn. I didn't found this topic in documentation. If
not, I would like to contribute on this topic.
Please guide me.
Thanks !!
------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id1051231&iu=/4140/ostg.clktrk
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
--001a1136bd706db3a204f17b5633
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable <div dir="ltr">There have been many people asking about contributing recommender systems to scikit-learn, and generally the response has been that it doesn&#39;t quite fit in with the library. Though it can be shoehorned somewhat perhaps, I recommend you take a look at �<a href="https://github.com/mendeley/mrec">https://github.com/mendeley/mrec</a>, which implements a number of recommender algorithms, depends in part on scikit-learn, and tries where possible to conform to the scikit-learn API.<div> <br></div><div>Nick</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Feb 3, 2014 at 3:23 AM, <a href="mailto:***@masonlive.gmu.edu">***@masonlive.gmu.edu</a> <span dir="ltr">&lt;<a href="mailto:***@masonlive.gmu.edu" target="_blank">***@masonlive.gmu.edu</a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">




<div>
<div style="direction:ltr;font-size:10pt;font-family:Tahoma">This is in response to the thread on recommender system implementation in scikit-learn. I would also like to know if any of the scikit-learn contributors are willing to mentor a
project which implements basic recommender system algorithms - collaborative filtering (user-based/item-based/matrix factorization) for Google Summer of Code. I feel the lack of a scalable, extensible open-source recommendation engine in python is an interesting
gap to fill and would like to try my hand at it during GSOC. There are a couple of interesting problems to address in this case like how to design a recommendation engine that conforms to the design of scikit-learn package as much as possible. Some of the
other challenges are implementing support for Sparse matrix operations. 
<div><span style="font-size:10pt"><br>
</span></div>
<div><span style="font-size:10pt">Thanks,</span></div>
<div><span style="font-size:10pt">Nikhil </span>
<div>
<div><br>
<div style="font-size:16px;font-family:Times New Roman">
<hr>
<div style="direction:ltr"><font face="Tahoma" color="#000000"><b>From:</b> MIT SHAH [<a href="mailto:***@gmail.com" target="_blank">***@gmail.com</a>]<br>
<b>Sent:</b> Sunday, February 02, 2014 9:39 AM<br>
<b>To:</b> <a href="mailto:scikit-learn-***@lists.sourceforge.net" target="_blank">scikit-learn-***@lists.sourceforge.net</a><br>
<b>Subject:</b> [Scikit-learn-general] Contributing in a New Topic : Recommender Systems<br>
</font><br>
</div>
<div></div>
<div>
<div dir="ltr">Hi,
<div><br>
</div>
<div>     I want to know whether there are algorithms on &quot;Recommender Systems&quot; in scikit-learn. I didn&#39;t found this topic in documentation. If not, I would like to contribute on this topic.</div>
<div>     Please guide me.</div>
<div><br>
</div>
<div>Thanks !!</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>

<br>------------------------------------------------------------------------------<br>
Managing the Performance of Cloud-Based Applications<br>
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.<br>
Read the Whitepaper.<br>
<a href="http://pubads.g.doubleclick.net/gampad/clk?id=121051231&amp;iu=/4140/ostg.clktrk" target="_blank">http://pubads.g.doubleclick.net/gampad/clk?id=121051231&amp;iu=/4140/ostg.clktrk</a><br>_______________________________________________<br>

Scikit-learn-general mailing list<br>
<a href="mailto:Scikit-learn-***@lists.sourceforge.net">Scikit-learn-***@lists.sourceforge.net</a><br>
<a href="https://lists.sourceforge.net/lists/listinfo/scikit-learn-general" target="_blank">https://lists.sourceforge.net/lists/listinfo/scikit-learn-general</a><br>
<br></blockquote></div><br></div>

--001a1136bd706db3a204f17b5633--
Continue reading on narkive:
Loading...