When social network has reached hundreds of million users, the analysis of data in social network services becomes very important. Understanding how nodes interconnect in large graphs is an essential problem in many fields. In order to find connecting nodes between two nodes or two groups of source nodes in huge graphs, we propose a parallelized data-mining algorithm to get the shortest path between nodes in a social network based on HBase distributed key/value store. Our algorithm can achieve the shortest path among different nodes in network under the parallel environment. We analyze the social network model by this algorithm first, and then optimize the output from cloud platform by using the intermediary degrees and degree central algorithm. Finally, with a simulated social network, we validate the efficiency of the proposed algorithm. The experiment results indicate that our algorithm can improve the efficiency of parallel breath-first search (BSF).
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Discrete Mathematics and Combinatorics
- Control and Optimization
- Computational Theory and Mathematics
- Applied Mathematics