Heterogeneous Information Network (HIN) is a natural and general representation of data in modern large commercial recommender systems which involve heterogeneous types of data. HIN based recommenders face two problems: how to represent the high-level semantics of recommendations and how to fuse the heterogeneous information to make recommendations. In this paper, we solve the two problems by €rst introducing the concept of meta-graph to HINbased recommendation, and then solving the information fusion problem with a “matrix factorization (MF) + factorization machine (FM)” approach. For the similarities generated by each meta-graph, we perform standard MF to generate latent features for both users and items. With di‚erent meta-graph based features, we propose to use FM with Group lasso (FMG) to automatically learn from the observed ratings to e‚ectively select useful meta-graph based features. Experimental results on two real-world datasets, Amazon and Yelp, show the e‚ectiveness of our approach compared to stateof-the-art FM and other HIN-based recommendation algorithms.