Mailing-List: contact user-help@predictionio.incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@predictionio.incubator.apache.org
MIME-Version: 1.0
From: Umesh Awasthi <umeshawasthi@gmail.com>
Date: Sun, 23 Oct 2016 20:08:03 -0700
Message-ID: <CAAJB1JMVhXDNVk=5_tDWdRUGC-Rqw34PL_tDhwBqfcKpUbe6KA@mail.gmail.com>
Subject: Apache Mahout or PredictionIo for recommendation engine
To: user@predictionio.incubator.apache.org
Content-Type: multipart/alternative; boundary=94eb2c07374e4a6855053f93b3f9
archived-at: Mon, 24 Oct 2016 03:08:10 -0000

--94eb2c07374e4a6855053f93b3f9
Content-Type: text/plain; charset=UTF-8

I am looking in to both Apache Mahout and PredictionIo to create a
recommendation engine, I checked both on high level and seems both are good
enough to support my use case with PredictionIo have some support for REST
Based API.

Below is a very high level overview about the proposed API

   1. Send customer data to ML server using Click-Stream API
   2. Send customer and order information to Server ( Either real time or
   batch)
   3. Provide recommendation to customer at real time using REST API.
   4. Plan is to deploy this engine as a service and ability to integrate
   with different platforms.

initially I was looking in to Apache Mahout as possible option but it seems
that PredictionIO fits more in to my use case and it is more focused for
such tasks.

Can some one help me in following points

   1. What is best option for me
   2. Which is best fit for deploying engine as a service.

Initially PredictionIO was based on Mahout but I think they have changed it
completely with latest release.

Not sure how subjective my question is but any help is much appreciated.


-- 
With Regards
Umesh Awasthi
http://www.travellingrants.com/

--94eb2c07374e4a6855053f93b3f9
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail-post-text">

<p>I am looking in to both Apache Mahout and PredictionIo to create a=20
recommendation engine, I checked both on high level and seems both are=20
good enough to support my use case with PredictionIo have some support=20
for REST Based API.</p>

<p>Below is a very high level overview about the proposed API</p>

<ol><li>Send customer data to ML server using Click-Stream API</li><li>Send=
 customer and order information to Server ( Either real time or batch)</li>=
<li>Provide recommendation to customer at real time using REST API.</li><li=
>Plan is to deploy this engine as a service and ability to integrate with d=
ifferent platforms.</li></ol>

<p>initially I was looking in to Apache Mahout as possible option but it
 seems that PredictionIO fits more in to my use case and it is more=20
focused for such tasks.</p>

<p>Can some one help me in following points</p>

<ol><li>What is best option for me</li><li>Which is best fit for deploying =
engine as a service.</li></ol>

<p>Initially PredictionIO was based on Mahout but I think they have changed=
 it completely with latest release.</p>

<p>Not sure how subjective my question is but any help is much appreciated.=
</p>
    </div><br clear=3D"all"><br>-- <br><div class=3D"gmail_signature">With =
Regards<br>Umesh Awasthi<br><a href=3D"http://www.travellingrants.com/" tar=
get=3D"_blank">http://www.travellingrants.com/</a><br><br>=C2=A0 =C2=A0 </d=
iv>
</div>

--94eb2c07374e4a6855053f93b3f9--