This paper proposes a method of generating synthetic data for book recommendation based on knowledge graph embedding (KGE). Collaborative filtering (CF) has been widely and commonly used in recommender systems, which determines items to recommend from a rating matrix. However, its performance significantly decreases when enough amount of rating data is unavailable. In addition, privacy concerns have become a major issue in recent years. To solve these problems, the approach of generating synthetic data based on the statistical characteristics of real data is attracting attention in the field of machine learning and recommendation. The proposed method constructs a knowledge graph from user’s reading history and synthesizes a rating matrix by simulating the user’s reading behavior based on the link prediction using KGE. The effectiveness of the proposed method is shown by comparing it with an actual rating matrix in terms of precision and recall in a book recommendation task.