〔Abstract〕 Objective Through the analysis of the diagnosis and treatment data of coronavirus disease 2019 (COVID-19) infected people in Shenzhen, the clinical characteristics of COVID-19 patients from January 2020 to March 2022 were compared to find out the changing rules of the disease conditions of COVID-19 patients. Methods The Python programming tool was used to preprocess the data of specific diseases, and 1,668 effective treatment records were selected for analysis. The χ² test method was used to compare the treatment data to find the differences. The K-means clustering model was analyzed and visually displayed to compare the clinical manifestations of different novel coronavirus strains. To summarize the pathogenic characteristics of Omicron strain of COVID-19 in Shenzhen since 2022. Results The transmission rate of Shenzhen 2022 COVID-19 Omicron BA.2 was fast, but the virulence was weak, the rate of severe illness was low, and the overall condition of the patients who relapse positive was relatively mild, which was lower than that of the classic COVID-19 strain in 2020, but the average number of days between relapse positive was shortened. Conclusion Regular data mining is used to track and analyze the diagnosis and treatment data of COVID-19 patients, which can discover the evolving rules of the disease of patients with different COVID-19 strains in different periods, providing scientific decision-making basis for the prevention and control of COVID-19. In addition, the number and rate of adolescent infection increased significantly, and most of the adolescents who recovered from the disease were infected again within 1 week, suggesting that this period is a susceptible period. Therefore, it is necessary to strengthen the publicity, improve the awareness of safety, and pay attention to virus protection.