Cooperative decision making algorithm for large networks using MapReduce programming model

Abstract

Recently, due to information explosion in our environment, complex networks appear more frequently. Many types of networks, such as social networks, biological structures, telecommunication networks and many others, consist of huge amount of data which grows exponentially. Obviously, there is a problem with processing such big datasets, especially in order to provide support for decisions making. In general, structure of networks is not independent. It means behaviour of single node in network depends on behaviour of others. According to this property, decision making for particular network's node must be based on gathered information from other nodes. Such a collaborative decision making, based on dependencies of network's components induce more accurate results. In this paper it is presented an early stage consideration of sampling techniques based on node degree for decision making in huge networks based on information sharing and propagation. We solve a problem of decision assignment for each node using information propagation in a network. Proposed collaborative decision making solution assumes decomposition of the problem that allows distributed computing. © 2012 Springer-Verlag.