BalanSiNG: Fast and Scalable Generation of Realistic Signed Networks


How can we efficiently generate large-scale signed networks following real-world properties? Due to its rich modeling capability of representing trust relations as positive and negative edges, signed networks have spurred much interests with various applications. Despite its importance, however, existing models for generating signed networks do not correctly reflect properties of real-world signed networks.

In this paper, we propose BalanSiNG (Balanced Signed Network Generator), a novel, scalable, and fully parallelizable method for generating large-scale signed networks following realistic properties. We identify a self-similar balanced structure observed from a real-world signed network, and simulate the self-similarity via Kronecker product. Then, we exploit noise and careful weighting of signs such that our resulting network obeys various properties of real-world signed networks. BalanSiNG is easily parallelizable, and we implement it using Spark. Extensive experiments show that BalanSiNG efficiently generates the most realistic signed networks satisfying various desired properties.


BalanSiNG is described in the following paper:


The codes used in the paper are available. [c++] [spark]


Name #Nodes #Edges Description Source Download
Epinions 131,828841,372 Trust social network in Epinions service TRUSTLET Link
BitcoinO 5,88135,592 Trust social network in Bitcoin-OTC service SNAP Link
BitcoinA 3,78324,186 Trust social network in Bitcoin-ALPHA service SNAP Link
Congress 219764 Politician network CONVOTE Link