The development of digital technology provides a new way for teachers to obtain, integrate and use ideological and political education resources, which can enrich the teaching content, stimulate students’ interest in learning, and thus improve the teaching effect. This paper proposes an ideological and political education resources integration platform based on Web service technology and J2EE technology, calculates the similarity of different types of ideological and political education resources using semantic association similarity, and constructs an educational resources information matching algorithm by combining metadata similarity. Then, the hash algorithm is used to pre-process the retrieval request of ideological and political education resources, and the unsupervised cross-modal hash retrieval model is established by combining the graph convolutional neural network, which is used to achieve the fast retrieval of ideological and political education resources. When the resource matching result set is upgraded from Top5 to Top50, the average search accuracy of the resource information matching algorithm always remains above 65%. The cross-modal hash retrieval model sees an average improvement of 2.93% and 1.52% on the tasks of retrieving images from texts and text from images. The resource integration platform for ideological and political education helps to achieve rapid retrieval and matching of ideological and political education resources and provides new support for diversified teaching of ideological and political education.