Abstract: The paper emphasizes on the correlativity content retrieving method that are studied to compare the incoming query demand of the user with the content provided by the content management system.Similarity measures are done in the content vector space and the correlativity are ranked during the process of the content retrieval.The process and the algorithm of the correlativity content retrieving methods are proposed and the validity of the algorithm is analyzed.The trained self-organization neural network is used to cluster the query demand and the matching work is just done in the classification the query belongs to.The policy of intelligent clustering based correlativity content retrieval can suggest the different users how correlative the content is to their query demands so that the users can quickly select the content they concerns.