Abstract: 【Purpose/significance】Scientific and technological report data is a fundamental strategic resource for the country, and it is im⁃ perative to research the technologies and methods for its development and utilization. By identifying the research themes in the field of biotechnology and their evolution, we can fill the gap in the development and application of scientific report data【. Method/process】A text corpus of scientific reports in the biotechnology field was constructed, and a BERTopic model was trained to explore and analyze the evo⁃ lution of research themes in the field.【Results/conclusion】A total of 30 themes in the biotechnology field were identified based on the BERTopic model. The hierarchical clustering of themes revealed nine major research directions in biotechnology: plant genomics and ge⁃ netic modification, genetic engineering and industrial biotechnology, the application of biotechnology in biological and ecological environ⁃ ments, veterinary virology and immunology, molecular genetics and biochemistry, cardiovascular metabolic health and neurobiology, bone biology and regenerative medicine, and biomedical and clinical research【. Innovation/limitation】The constructed model can better iden⁃ tify the research themes presented in scientific report data, and the quality of the generated thematic descriptions in the field of biotech⁃ nology is satisfactory. The corpus performs semantic analysis on the abstracts and time fields in the scientific report data but does not ana⁃ lyze other fields.