Thanks to IT's meteoric rise to prominence, it's now permeating every facet of modern life. Now more than ever, information instruction is all the rage, particularly at the university level. Much progress has been achieved in the creation of informatization teaching in higher education, which is now in the implementation stage. An essential change that may better meet the high standards of higher education in modern society is the building of information teaching in colleges and universities. The challenge of information teaching construction in higher education cannot be addressed by conventional methods of instruction. As a result, this research suggests and evaluates a deep learning-based information teaching structure. In order to minimize interference elements in information teaching construction, computer technology is used to develop it. Indicators are then segregated according to the needs of information teaching construction. The next step is for computer technology to assess the outcomes of informatization teaching in colleges and universities thoroughly before developing an information teaching construction plan based on those outcomes. According to the results of a MATLAB simulation, when compared to conventional methods of instruction, the practicality and dependability of information teaching based on deep learning are superior in higher education.