Abstract In response to the current information needs of higher vocational mathematics and cultural teaching, Hadoop framework and collaborative recommendation algorithms are used to apply these two technologies to the construction of a shared platform for higher vocational mathematics and cultural infiltration teaching. Then through experiment and comparison, let the same teacher conduct mathematics and cultural infiltration teaching in two different classes. The results show that by investigating higher vocational students’ interest in mathematics and their understanding of mathematics, the number of people interested in vocational mathematics classes increased by 14.6%, and the number of people interested in control classes only increased by 10%. At the same time, in the assessment of mathematics culture penetration knowledge points, the number of outstanding people in the experimental group was significantly more than that in the control group.