This work focuses on recommending businesses to an user retrieved
from Yelp Dataset. The first part focuses on manipulating data to
calculate the review rating in a more clever way. The collaborative filtering
based on user similarity was later implemented. In the final section the UV Decomposition phase was carried out: the performance of this dimensional
reduction approach (written from scratch) was compared with that of the
spark.mllib.ALS through training and testing phases.
- Basic Recommendation
- Collaborative Filtering
- UV Decomposition
- ALS