Inclusive robots are capable of natural interaction with the working environment, people, and other robots, autonomously adapting to complex dynamic environments and collaborating in operations. The sensitive and considerate natural interaction is one of the research hotspots for inclusive service robots, and there is an urgent need in the industry for inclusive robots to possess the ability to understand complex semantic information. This book is based on the field of information and knowledge extraction using deep learning methods. Starting from learning text representations, it systematically introduces methods for acquiring named entities and entity relationships in real-world knowledge information and delves into achieving robust entity relationship analysis in open domains.