Large-scale social network analysis based on MapReduce

Social Networks Service(SNS), is becoming more and more popular and a lot of studies have been carried out in this active field. However, traditional analysis methods based on single machines is not suitable because the network is growing too large. MapReduce, a progromming paradigm proposed by Google, gives us a new approach to solve large-scale social networks analysis problem by making use of the power of multi-machines. In this paper, we use Hadoop - an open source implementation of MapReduce - to conduct a series of analyses on large-scale social networks including several distributions, clustering coefficient and diameter. We believe that our work will shed light on the study of measuring large-scale social networks. ? 2010 IEEE.EI