摘要

For the existing problems in the traditional way of mining frequent Close-Graph, a new approach based on top-down mining strategy is proposed in this paper. In order to reduce the unnecessary operation and the storage of redundant graphs in the mining process, firstly, this approach simplifies judgment isomorphism through the necessity of graph isomorphism;secondly, the symmetry of graph is used to relate the graph sets;finally, the efficiency of algorithm is improved by constructing the tree structure using graph isomorphism determination, and completing the calculation of support without judging subgraph isomorphism. Experiments show that the algorithm is superior to the existing CloseGraph mining algorithms.

  • 出版日期2012

全文