hadoop-mapreduce-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Marcos Ortiz <mlor...@uci.cu>
Subject Re: How can I build a Collaborative Filtering recommendation framework based on mapreduce
Date Sat, 31 Mar 2012 16:50:37 GMT
Mahout is built precisely, so I think that you can evaluate it again.
It has to two collaborating filtering algorithms:

- Non-distributed recommenders ("Taste")
https://cwiki.apache.org/confluence/display/MAHOUT/Recommender+Documentation

- Distributed recommenders ("Item-based")
https://cwiki.apache.org/confluence/display/MAHOUT/Itembased+Collaborative+Filtering

- First-time FAQSs
https://cwiki.apache.org/confluence/display/MAHOUT/Recommender+First-Timer+FAQ

About the test that you did with Mahout:
- Which are the features of your machine?
If you are working with 175M of data, a single machine
is not the best way to do it. It's more worthy if you use
small Hadoop cluster for this (1 NN/JT and 3 DN/TT), and then
you can ask on the Mahout mailing list how to improve the performance
of your system.

Regards

On 3/31/2012 6:17 AM, chao yin wrote:
> Hi all:
> I'm new to mapreduce, but familiar with Collaborative Filtering 
> recommendation framework.
> I tried to use mahout to do this work. But it disappointed me. My 
> machine work all day to do this job without any result with about 175M data.
> Is there anyone knows anything about Collaborative Filtering 
> recommendation framework based on mapreduce, or mahout, any suggestion 
> to improve performance?
> 
> -- 
> Best regards,
> Yin

-- 
Marcos Luis Ortíz Valmaseda (@marcosluis2186)
 Data Engineer at UCI
 http://marcosluis2186.posterous.com

10mo. ANIVERSARIO DE LA CREACION DE LA UNIVERSIDAD DE LAS CIENCIAS INFORMATICAS...
CONECTADOS AL FUTURO, CONECTADOS A LA REVOLUCION

http://www.uci.cu
http://www.facebook.com/universidad.uci
http://www.flickr.com/photos/universidad_uci

Mime
View raw message