With the rapid development of modern basketball in the process of high popularity and professional development, fierce confrontation and high-level competition make people pay special attention to the selection and cultivation of outstanding talents, among which the quality and level of youth training will play a key role in the future competition. Good ecological environment of basketball class provides basic conditions for teachers and students to carry out basketball teaching activities, and promotes the improvement of students' basketball technical level. By using data mining (DM) technology, this paper extracts implicit and useful knowledge from sports talent data, explores the development and management rules of college basketball talent training, and then explores a new mode of sports basketball professional talent training under the influence of ecological environment. This will enable the educational ecology of basketball teaching to be used and effectively promoted, and achieve sustainable development. The results show that the accuracy of sports talent information mining algorithm in this paper is improved by 23.85% compared with traditional methods. Universities should change the traditional teaching mode of basketball professionals, and cultivate a number of basketball professionals who meet the needs of the society through the structure of strategic education and the countermeasures of comprehensive quality and ability training.