This paper presents a graph convolutional network (GCN)-based collaborative filtering model considering personality bias. Collaborative filtering (CF) can cause the personality bias that is the performance difference depending on the users’ personality. This indicates that some users negatively affect the recommendation performance. We quantity the user’s personality with the OCEAN model to deal with each user differently according to their personality in a GCN-based CF model. By considering the personality bias, this negative effect can be reduced. Experiments demonstrated the efficacy of the proposed model.