Open AccessMSP-N: Multiple selection procedure with ‘N’ possible growth mechanisms
Author(s) -
Pradumn Kumar Pandey,
Mayank Singh
Publication year - 2019
Publication title -
plos one
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.99
H-Index - 332
ISSN - 1932-6203
DOI - 10.1371/journal.pone.0224383
Subject(s) - node (physics) , degree distribution , exponential growth , complex network , spectral radius , computer science , exponential function , degree (music) , preferential attachment , cluster analysis , task (project management) , radius , selection (genetic algorithm) , clustering coefficient , distribution (mathematics) , scale free network , evolving networks , mathematics , artificial intelligence , physics , computer network , engineering , mathematical analysis , eigenvalues and eigenvectors , quantum mechanics , world wide web , acoustics , systems engineering
Network modeling is a challenging task due to non-trivial evolution dynamics. We introduce multiple-selection-procedure with ‘N’ possible growth mechanisms ( MSP-N ). In MSP-N , an incoming node chooses a single option among N available options to link to pre-existing nodes. Some of the potential options, in case of social networks, can be standard preferential or random attachment and node aging or fitness. In this paper, we discuss a specific case, MSP-2 , and shows its efficacy in reconstructing several non-trivial characteristic properties of social networks, including networks with power-law degree distribution, power-law with an exponential decay (exponential cut-off), and exponential degree distributions. We evaluate the proposed evolution mechanism over two real-world networks and observe that the generated networks highly resembles the degree distribution of the real-world networks. Besides, several other network properties such as high clustering and triangle count, low spectral radius, and community structure, of the generated networks are significantly closer to the real-world networks.