Examples for Recommendation

[srinivasaraov]

Hi,

I was wondering if you have any examples of recommendation systems using your code?

How can we build a recommendation model and predict using MatrixFactorization?

[asmirn1]

Hello,

I'm using this NMF code for solving the blind source separation problem. You can find details of the problem I was working on here: https://pubs.acs.org/doi/10.1021/acs.analchem.9b01424

I didn't try using this code for a recommendation system, although it should be possible. So, if you have any specific problem in mind, I can help with building a recommendation system for it with this code.

If you are interested, this article gives some basic examples of using NMF in recommendation systems:
https://medium.com/logicai/non-negative-matrix-factorization-for-recommendation-systems-985ca8d5c16c

[srinivasaraov]

Thanks @asmirn1

I am trying to use Weka with Matrix Factorization for Recommendation with no luck. I am trying to build a collaborative filtering based recommendation using NMF (In Java) for movie reviews(movie lens data) to begin with. Goal is to Recommend movies to a user.

[asmirn1]

Hi @srinivasaraov,

I've read a little more about the collaborative filtering with NMF. It looks like NMF should be able to work with sparse matrices in order to use it in a recommendation system. The algorithms that I've developed currently work only with dense matrices. So I doubt that you could use JavaNMF for it.

However, if you are interested in implementing the stochastic gradient descent algorithm for performing NMF on sparse matrices, I will gladly help to develop it and add it to JavaNMF.