The application of link prediction in web recommendation systems

The link prediction in complex networks is about how to predict the possibility about the connection of unknown nodes via analyzing known nodes inside the network. Considering the current needs of the education websites as the starting point, this paper introduces the specific content of the link prediction based on detailed link-node similarity prediction methods. And this paper describes the measurement accuracy of the node similarity. This provides a guideline for the development of web recommendation systems.