XGCN: a library for large-scale graph neural network recommendations

Xiran Song, Hong Huang, Jianxun Lian, Hai Jin

Abstract: This work introduces a GNN library, XGCN, which is designed to assist users in rapidly developing and running large-scale GNN recommendation models. We offer highly scalable GNN reproductions and include a recently proposed GNN model: xGCN. Experimental evaluations on datasets of varying scales demonstrate the superior scalability of our XGCN library.