The era of big data has brought about a new approach to assessing universities and institutions. This paper delves into university algorithm courses’ ideological and political construction by applying big data analysis techniques and data mining theories. This exploration can effectively promote the reform of university teaching evaluation, enhancing higher education’s overall ideological and political education level. This paper discusses the ideological and political construction of algorithm courses and teaching practices under the big data mining framework. The results show that the accuracy of big data mining is significantly higher than traditional fuzzy evaluation models. After optimizing the hidden layer nodes of the FNN, the accuracy of the evaluation network remains almost unchanged. Therefore, the evaluation model based on the FNN neural network is more practical. This result contributes to improving teaching quality and effectiveness in algorithm courses.