mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Winnu Ayi Satria <winnu...@gmail.com>
Subject Re: MongoDBDataModel doesn't work?
Date Wed, 01 Aug 2012 21:06:10 GMT
After checking through debugger, I could confirm using the simple code from
Mahout In Action book and MongoDBDataModel, it works. Actually it is
trivial problem, the actual userID in MongoDB or CSV file is different with
userID inside MongoDBDataModel. So is the itemID.

for example:

recommendation:RecommendedItem[item:*5*, value:4.257081]
recommendation:RecommendedItem[item:*10*, value:4.0]

itemID has been changed. not [101-107] anymore.

Using sample code from the internet link I pasted before must be handled
with thoroughly while it looks like show no similarity (similarity=NaN)
among the users and (SVDRecommender might be the appropriate
solution). That iss why I totally lost.

Okay, thanks for the suggestion. :)

On Wed, Aug 1, 2012 at 9:43 PM, Sean Owen <srowen@gmail.com> wrote:

> If the data is 'really' there in the DataModel you seem to have ruled
> out all the differences. ;) I imagine there is something slightly
> amiss. Can you step through with a debugger to see what the
> UserSimilarity calculates? look what data it gets and see if it makes
> sense. If it seems to, double-check against FileDataModel. If you can
> say more about the difference maybe I can offer a better guess.
>
> The Mongo-backed class is a bit of a toy. It loads all the data in
> memory. While that's handy just insofar as it lets you consume data
> that happens to be in Mongo, it's not any more scalable than something
> backed by a file.
>
> Anything real-time has to be in memory, really.  Anything not
> real-time -- doesn't matter where your data is put since you're likely
> distributing with something like Hadoop, which consumes data in a
> linear and leisurely way.
>
>
> On Wed, Aug 1, 2012 at 8:37 PM, Winnu Ayi Satria <winnuayi@gmail.com>
> wrote:
> > Hi all,
> >
> > I am trying to combine MongoDB and Mahout using the same code in Mahout
> In
> > Action book, chapter 2. The very first code. But now I replaced the
> source,
> > user-item-preference, not from CSV file but from MongoDB. So the model is
> > instantiated from MongoDBDataModel, not FileDataModel anymore. The CSV
> file
> > has been copied to MongoDB. Everything is same, except how to create the
> > model.
> >
> > However, it doesn't work. The UserSimilarity class can not generate
> > similarity table from the data model. I checked the data inside the
> > instantiated model, and yes, all users, items, and preference do exist.
> Is
> > there something problem with MongoDBDataModel and UserSimilarity?
> >
> > I also tried the sample from here,
> > http://www.smartlab.at/articles/recommendation-engine/, that uses
> > SVDRecommender and it works.
> >
> > Too bad, not so many example or reference about MongoDBDataModel in
> > internet while combining Mahout and MongoDB would be interesting.
> >
> > Anyone has ever experienced with this problem?
> >
> > --
> > Best regards,
> >
> > Winnu Ayi Satria
>



-- 
Best regards,

Winnu Ayi Satria

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message