Your question is not clear. In GraphLab Create we support multiple recommender models.
Matrix Factorization based is one family of algorithms, and is explained here: https://dato.com/products/create/docs/generated/graphlab.recommender.factorizationrecommender.FactorizationRecommender.html
Item based methods are explained here: https://dato.com/products/create/docs/generated/graphlab.recommender.itemsimilarityrecommender.create.html#graphlab.recommender.itemsimilarity_recommender.create
You are encouraged to read the explanations along with the reference papers. Let us know if you have further questions.