The Internet of Things based on data mining algorithms has begun to be used in the early warning of dynamic systems.in recent years, the diversified application of data has developed rapidly, especially in data mining. With the help of data mining, we can select valuable data for research and analysis in huge data. It can be said that data mining is active in all walks of life. This paper focuses on the application of data mining algorithms in the internal logic and dynamic early warning system of sports ideological and political education. Sports ideological and political education plays an important role in colleges and universities. Sports ideological and political education usually includes logical elements such as sportsmanship cultivation, teamwork guidance, competitive ethics, and sports psychology counseling. Among them, the data is huge and mixed. If the effective data cannot be fully screened, it will not only bring defects to the work, but also have a negative impact on the management of the school. This paper investigates the topic of the internal logic of political and psychological learning with the help of data extraction and clustering-based analysis. Based on the data of assessment quantification table, this paper divides the obtained data into four attributes. It calculates that the scores of the four attributes are 0.6171, 0.5927, 0.536, and 0.5917 respectively, and then it is concluded that the management attitude of the counselors in the work assessment is at a high level. But there are certain problems with the management method. Combined with the actual situation, this paper sets early warning thresholds for different attributes and their index scores, and conducts timely early warning analysis on the related work of thinking and political study. Meanwhile, this paper also randomly selects 1000 assessment forms, using traditional methods and dynamic early warning systems respectively. It is concluded that the accuracy rate, time efficiency ratio and reliability performance ratio of the early warning system are 0.25, 0.3 and 0.45 higher than those of the traditional analysis method, respectively. Data mining under the Internet of Things has some important reference significance for the development of artificial intelligence.
