Graph-structured data is ubiquitous throughout the natural and social sciences, from telecommunication networks to quantum chemistry. Building relational inductive biases into deep learning architectures is crucial for creating systems that can learn, reason, and generalize from this kind of data. Recent years have seen a surge in research on graph representation learning, including techniques for deep graph embeddings, generalizations of convolutional neural networks to graph-structured data, and neural message-passing approaches inspired by belief propagation. These advances in graph representation learning have led to new state-of-the-art results in numerous domains, including chemical synthesis, 3D vision, recommender systems, question answering, and social network analysis. This book provides a synthesis and overview of graph representation learning. It begins with a discussion of the goals of graph representation learning as well as key methodological foundations in graph theory and network analysis. Following this, the book introduces and reviews methods for learning node embeddings, including random-walk-based methods and applications to knowledge graphs. It then provides a technical synthesis and introduction to the highly successful graph neural network (GNN) formalism, which has become a dominant and fast-growing paradigm for deep learning with graph data. The book concludes with a synthesis of recent advancements in deep generative models for graphs—a nascent but quickly growing subset of graph representation learning.

We met again in front of the statue of Gottfried Wilhelm von Leibniz in the city of Leipzig. Leibniz, a famous son of Leipzig, planned automatic logical inference using symbolic computation, aimed to collate all human knowledge. Today, artificial intelligence deals with large amounts of data and knowledge and finds new information using machine learning and data mining. Machine learning and data mining are irreplaceable subjects and tools for the theory of pattern recognition and in applications of pattern recognition such as bioinformatics and data retrieval. This was the fourth edition of MLDM in Pattern Recognition which is the main event of Technical Committee 17 of the International Association for Pattern Recognition; it started out as a workshop and continued as a conference in 2003. Today, there are many international meetings which are titled “machine learning” and “data mining”, whose topics are text mining, knowledge discovery, and applications. This meeting from the first focused on aspects of machine learning and data mining in pattern recognition problems. We planned to reorganize classical and well-established pattern recognition paradigms from the viewpoints of machine learning and data mining. Though it was a challenging program in the late 1990s, the idea has inspired new starting points in pattern recognition and effects in other areas such as cognitive computer vision.