Aiming at the problem of sampling risk in traditional text statistics, text mining technology is used to extract information with analytical value in relevant literature, natural language processing technology is introduced to extract and clean literature information for subject words, and text clustering is used to combine knowledge graph and data visualization. Provide the basis for analyzing the current situation, hotspots and development trends of related literature research directions. Taking education literature as an example, the research shows that natural language processing technology and clustering method combined with data visualization can provide strong data support for analyzing the current research hotspots and development trends in the field of education.