In the era of highly developed Internet technology, the learning resources on the Internet show an exponential growth trend. With the diversification and disorder between various learning objects and concepts, the recognition of the dependencies between them will have a major impact on computer education. Aiming at the solution to this problem, this paper proposed a concept dependency recognition method for Wikipedia. Using the characteristics of the concept in Wikipedia, a set of recognition concept dependency model was designed, and the machine learning based classification algorithm was used to test on the public data set. The experimental results show that with high accuracy and recall rate, the model can effectively discover the dependencies between concepts.