Knowledge graph reasoning involves inferring new knowledge from existing data in a knowledge graph. Theoretical aspects focus on understanding the complexity and tractability of reasoning tasks. The book explores parallel tractability, identifying datalog programs with reasoning bounded by NC complexity, ensuring parallelism. Practical aspects involve improving reasoning efficiency and scalability. The book proposes using Boolean circuits to construct neural networks for reasoning, enhancing parallelism and efficiency. It also discusses the application of these theoretical results in creating parallel tractable knowledge graphs and constructing logic neural networks, demonstrating the potential for practical applications in various domains.