Sparsity Data Reduction In Textual Network Analysis. An exercise on Sustainability meaning

In this paper, we propose a new strategy to derive an unweighted adjacency matrix from an affiliation matrix. The strategy is based on the use of a biclustering technique in order to reduce the sparsity of the matrix without changing the network structure. As an example, we implemented this approach to seek the common meaning of the term sustainability by using an affiliation matrix characterized by a core-periphery structure. The application of BiMax Biclustering algorithm shows a sparsity reduction of the unweighted adjacency matrix with an invariant network structure.