Due to the recent development of data analysis techniques, technologies for detecting communities through information expressed in social networks have been developed. Although it has several advantages, including the ability to effectively share recommended items through an estimated community, it may cause personal privacy issues. Therefore, recently, the problem of hiding the real community well is being studied at the same time. As an example, Mittal \etal, proposed an algorithm called NEURAL that can hide this community well based on a metric called permanence. Based on this, in this study, we propose a Reverse NEURAL (R-NEURAL) algorithm that restores the community as well as the original graph structure using permanence. The proposed algorithm includes a method for well restoring not only the community hidden by the NEURAL algorithm but also the original graph structure modified by recovered edges. We conduct experiments on real-world graphs and found that the proposed algorithm recovers well the hidden community as well as the graph structure.