Abstract: This paper systematically reviews the latest research developments in Named Entity Recognition (NER) technology based on deep learning. The article first outlines the background and significance of NER technology, followed by an analysis of NER methods based on early deep learning architectures, including convolutional neural networks, recurrent neural networks, and graph neural networks. Next, it focuses on the NER technology route based on Transformer models, particularly the practical applications of BERT and its derivative variants in NER. Additionally, the paper explores the current research status and challenges faced by large language models in NER technology. Finally, it discusses future research directions for NER technology and related technical branches, and summarizes the contributions of this study, aiming to provide researchers and practitioners in the field with a comprehensive overview of the application of deep learning in NER.