This study focuses on the innovation of college education and teaching based on big data mining and explores how to improve the quality and efficiency of education and teaching through big data technology. With the rapid development of information technology, the application of big data in the field of education has gradually become an effective means to improve education management and teaching mode. At present, in the reform of education and teaching, the traditional teaching methods have been difficult to meet the individualized and diversified educational needs. Through big data mining and machine learning algorithms, this study attempts to build a new data-driven education and teaching innovation path. Methods such as data collection, data preprocessing, model construction, and optimization are adopted to build an educational innovation model based on big data in combination with the actual education in colleges and universities. Through data collection and analysis, design and optimization of education model, and evaluation of the effect of education and teaching reform, this paper analyzes the educational data of colleges and universities, and puts forward strategies for personalized teaching and educational resource optimization. The results show that data-driven education model has advantages in improving students’ learning effect and teachers’ teaching efficiency. With the rational optimization of personalized teaching paths and teaching resources, the quality of education has been effectively improved. It also proves that big data technology can provide scientific basis for educational decision-making and promote educational innovation. This study provides the theoretical basis and practical path for data-driven innovation in the field of education, promotes the reform of education and teaching mode, and provides a new idea for the future research of educational technology.