Multi-fuzzy-objective graph pattern matching in big graph environments with reliability, trust and social relationship

Abstract

With the advent of the era of big data, the scale of data has grown dramatically, and there is a close correlation between massive multi-source heterogeneous data, which can be visually depicted by a big graph. Big graph, especially from Web data, social networks, or biometric data, has attracted more and more attention from researchers, which usually contains complex relationships and multiple attributes. How to perform efficient query and matching on big graph data is the basic problem on analyzing big graph. Using multi-constrained graph pattern matching, we can design patterns that meet our specific requirements, and find matched subgraphs to locate the required patterns to accomplish specific tasks. So how to find matched subgraphs with good attributes in big graph becomes the key problem on big graph research. Considering the possibility that a node in a subgraph may fail due to reliability, in order to select more and better matched subgraphs, in this paper, we introduce fuzziness and reliability into multi-objective graph pattern matching, and then use a multi-objective genetic algorithm NSGA-II to find the subgraphs with higher reliability and better attributes including social trust and social relationship. Finally, a reliability-based multi-fuzzy-objective graph pattern matching method (named as RMFO-GPM) is proposed. The experimental results on real data sets show the effectiveness of the proposed RMFO-GPM method comparing with other state-of-art methods.