Blog Recommender Based on Hypergraph Modeling Clustering Algorithm

In this paper, we presents a new blog recommender based on the hyper graph modeling clustering algorithm. The algorithm of this system is called multilevel clustering segmentation algorithm based on the hyper graph model. Under the hyper graph model, each blog will be abstracted as a vertex of a hyper graph, hyper edges will be built among the blogs which belong to the same result set gotten through the Google blog searching by one keyword. Then the problem of recommender is translated into the clustering of a hyper graph. Our multilevel clustering algorithm is then used to do the segmentation. And we set a new optimization index for the segmentation of a hyper graph, which is called modularity. Through the definition of modularity, it is not necessary for us to preset the partitioning number any more. We do the segmentation with the aim to achieve the optimum value of modularity. In addition, hyper graph is born with the ability to solve problems with high dimension data, and this ability, of course, will lead to more reasonable results compared with other models.