Graph representation learning goes beyond traditional data such as text, audio, and video by emphasizing the Euclidean features of the aforementioned objects and the relationships between them. This seminar will be dedicated to this type of data, and we will provide a three-session crash course on the basic knowledge of knowledge graphs, graph theory, as well as graph spectral theory. You will also have the opportunity to read the latest technical reports and research papers with your colleagues, including machine learning algorithms related to large language models, knowledge embedding, and social attribute prediction.