Uses for recommendation engines

Créer des moteurs de recommandation avec PySpark

Jamen Long

Data Scientist at Nike

blank matrix with users on y axis and movies on X axis

blank user/movie matrix and two factor matrices

matrix axes in factor matrices

factor matrices with latent feature axes

matrix and factor matrices with values

high horror movie ratings in matrix

low drama movie ratings in matrix

high drama ratings in matrix

low horror ratings in matrix

respective scores in factor matrices

original matrix with factor matrices with values

factor matrix with movies all scoring high in one latent feature

four unrelated movies with high scores in one latent feature

matrix with these movies with high avg scores

unrelated movies with respective Shakespeare work

Let's practice!

Créer des moteurs de recommandation avec PySpark