The availability of information about complex networks is severely restrained by issues, such as confidentiality and privacy. This poses a problem when analysing properties of networks that are of...Show moreThe availability of information about complex networks is severely restrained by issues, such as confidentiality and privacy. This poses a problem when analysing properties of networks that are of relevance to the general public. One example is the study of the resilience of banking networks to financial distress. We discuss a random graph model that reconstructs such unavailable networks, based on information that is either node specific or specific to groups of nodes. The procedure is determined by enforcing renormalizability, i.e. consistency when modelling renormalizations of networks. Then we propose a weighted semi-renormalizable extension of this model. Both models are tested on an empirical trade network, by analysing how well they captures properties that are commonly used to characterize networks. Their performances are shown to closely resemble that of the (weighted) fitness-induced Configuration Model.Show less
