Research On Graph Pattern Matching With Big Graph Data

Big graph,especially social networks,has attracted more and more scholars’ attention.One of the most important research fields is graph pattern matching,especially in big graph.Using graph pattern matching,we can design a pattern graph to meet special personalized requirements,and find the satisfied subgraphs.So how to effectively find better attributes in big graph in specific areas becomes a key issue in analyzing and processing big graph data.In order to quickly find matched subgraphs in big graph data and find better matched subgraphs in matched subgraphs set,a novel reliability-based multi-fuzzy-objective graph pattern matching is proposed in this thesis.The main work of this thesis is as follows:(1)In the existing networked multi-label classification,although only the labels of some unknown nodes are required to be determined,the labels of all nodes in the network must be inferred.This thesis proposes a multi-target core network consisting of the nodes with known label information,the target label nodes and the shortest path between them.(2)In the traditional multi-constrained graph pattern matching,using accurate numerical values to judge the degree of social trust with subjective consciousness and selecting more and better matched subgraphs,this thesis introduces fuzzy numbers,and proposes a method of multi-fuzzy-objective graph pattern matching.The experimental results prove that the proposed method can choose more and better matched subgraphs.(3)In view of the possible failure of a node in the matched subgraph in the traditional multi-constrained graph pattern matching,the reliability theory is introduced to evaluate the reliability of a matched subgraph,a reliability-based multi-fuzzy-objective graph pattern matching is proposed.The experimental results show that the proposed method is effective.(4)Aiming at the problem of how to find multi-objective are optimal matched subgraphs from numerous matched subgraphs,a multi-objective optimization algorithm is introduced in this thesis.