RENC: Recursive Estimation of Node Characteristics using topological structure of complex networks

Abstract

In this paper, we propose a recursive estimation method of node characteristics called RENC (recursive estimation of node characteristics) using the topological structure of a complex networks. RENC reduces the effect of noise by recursively estimating node characteristics using the topological structure of a network. In this paper, we also propose a network generation model called LRE (linkage with relative evaluation). The network generation model LRE is for simulating a social network, in which every node is likely to make decision based on relative evaluation, so that it can reproduce several characteristics of a social network. In this paper, we evaluate the effectiveness of our recursive estimation method RENC by applying RENC to several networks generated with LRE. Consequently, we show that the estimation accuracy of node characteristics can be improved by using our recursive estimation method RENC.