I have a small dataset of around 2 GB which has to be used to build a group based recommender system using Python 3.5 or above.
The choice of algorithm and methods is very specific. Time is limited.
Algorithm to be deployed is Non-negative Matrix Factorization. Aggregation strategies to be deployed are average, mean, median, Multiplicative Utilitarian, Least Misery and Most Pleasure.
Architecture is ready, you just have to implement and evaluate.
Major part is to work on how to convert the data from likes, shares, reading times, clickthroughs into explicit normalised ratings.
The project is short and easy. Someone who has previously worked on recommender systems and can finish the project ASAP is needed.
Further details shall be provided upon selection.
10 freelancers are bidding on average $206 for this job
Hi , I have experience of working on implicit feedback systems(non negative matrix factorization) for recommender systems. I can get this implemented in no time.