Subordinate Relationship Discovery Method Based on Directed Link Prediction

摘要

The subordinate relationship is the important relationship of the users in an enterprise. However, traditional knowledge discovery method cannot found this relationship. The directed link prediction can get the direction information of the nodes, and this direction information also reflect some subordinate relationship. In this paper, we propose a directed link prediction method to get the potential direction relationship in a network and judging the relationship between users in the network through a directed connection. Because the subordinate relationship cannot get directly, so we use the relationship recurrence rate to verify the effectiveness. The experiment proves that the prosed directed link prediction method can discover the relationship of users, and there is a stable relationship between users.