How to implement a physics engine?

What is the best framework to implement recommendation engine?

  • Edit:   A better way to form this question is. If you were to implement a professional grade recommendation engine.. something which is "good enough" for a production based environment, then how will your framework look like.. Backend, database?? distributed vs parallelism?? task vs data parallelism.. with just one restriction: No Java and no mahout :) ---------------------------------------------- Excluding Java ( no hadoop, mahout stuff... I think the mahout library is sub-optimal from machine learning point of view). How would you implement a recommendation engine for Netflix or say amazon.. C++ .. python.. or maybe functional programming (but no no java.. ) What is better data parallelism (map reduce way) or task parallelism (CUDA, MPI???????) Thanks.

  • Answer:

    The question boils down to, what clustering algorithms scale Unless you are using Random Projection , which scales linearly, you have a big task on your hand You should research scalable implementations of SVD, NMF, LDA, etc 2 good sources are   Gensim   GraphLab

Charles H Martin at Quora Visit the source

Was this solution helpful to you?

Other answers

Easiest answer I've contributed. You're looking for Neo4j. Open source match making app. Easily adaptable to serve as a recommendation engine based on tag attributes of a content object. http://maxdemarzi.com/2013/04/19/match-making-with-neo4j/ http://neolove.herokuapp.com/ http://www.neo4j.org/ It has a REST API, so pick your language, get the drivers, and jack-in Matrix style.

Kenny Bastani

The options will depend upon the volume of training data and the recommendation SLA. You need to ask question around then scale, is the recommendations going to be real-time or batch processed recommendation is the way to go etc and then try out a couple of frameworks. If you want to learn then you can try python based implementation, Mahout too has solid implementations.

Chandramani Tiwary

Programming Collective Intelligence has a python implementation in the book http://shop.oreilly.com/product/9780596529321.do

Niraj Juneja

Related Q & A:

Just Added Q & A:

Find solution

For every problem there is a solution! Proved by Solucija.

  • Got an issue and looking for advice?

  • Ask Solucija to search every corner of the Web for help.

  • Get workable solutions and helpful tips in a moment.

Just ask Solucija about an issue you face and immediately get a list of ready solutions, answers and tips from other Internet users. We always provide the most suitable and complete answer to your question at the top, along with a few good alternatives below.