The current knowledge mapping relationship optimization methods cannot obtain high-precision information. An optimization method of knowledge mapping relationship based on improved ant colony algorithm is proposed. The high-precision information of the network is obtained by using the cyclic network. The SGP problem is used to replace the optimization problem of the knowledge map relationship. The optimization objective function of the knowledge map relationship is constructed and solved by the improved ant colony algorithm. The optimization of the knowledge map relationship is realized. Experimental results show that the proposed method has high average accuracy, high knowledge accuracy and high knowledge coverage.