This review article systematically examines the advancements in drug-drug interaction (DDI) prediction using deep learning and knowledge graph techniques. The authors summarize available biomedical data and public databases related to drugs, discuss existing DDI prediction methods, and categorize them into three main classes: deep learning-based methods, knowledge graph-based methods, and combined deep learning with knowledge graph methods. They analyze the commonly used drug-related data and various DDI prediction methods, comparing them on benchmark datasets. The review also highlights challenges such as asymmetric DDI prediction and high-order DDI prediction. The article provides a comprehensive analysis of the current state of DDI prediction, emphasizing the importance of integrating deep learning and knowledge graph techniques to enhance prediction accuracy and interpretability.This review article systematically examines the advancements in drug-drug interaction (DDI) prediction using deep learning and knowledge graph techniques. The authors summarize available biomedical data and public databases related to drugs, discuss existing DDI prediction methods, and categorize them into three main classes: deep learning-based methods, knowledge graph-based methods, and combined deep learning with knowledge graph methods. They analyze the commonly used drug-related data and various DDI prediction methods, comparing them on benchmark datasets. The review also highlights challenges such as asymmetric DDI prediction and high-order DDI prediction. The article provides a comprehensive analysis of the current state of DDI prediction, emphasizing the importance of integrating deep learning and knowledge graph techniques to enhance prediction accuracy and interpretability.