In order to improve the teaching quality of vocational education, explore the vocational education model that meets the market demand, and build a distinctive vocational education system, it plays an important role in analyzing the research trend of vocational education. The process of text mining is first described in this paper, which primarily involves text data acquisition, text preprocessing, latent information extraction, and data statistical analysis. Then, in view of the shortcomings of the traditional K-means clustering algorithm, the algorithm is improved on this basis, and DT-Kmeans clustering is used to study and verify the effectiveness of the algorithm. The training tendency of each vocational education school was investigated. The keyword feature extraction was completed, and the integrated education model of each vocational education school was relatively unified. Among the top five keywords, each school had a high bias towards “natural science” and “ideological and political education”. Eighteen schools (90%) believe that “ideological and political education” occupies a “very important” position in student training. Two schools (10 percent) considered it “relatively important”.