Entropy-based randomisation of rating networks

Abstract

In the last years, due to the great diffusion of e-commerce, online rating platforms quickly became a common tool for purchase recommendations. However, instruments for their analysis did not evolve at the same speed. Indeed, interesting information about users' habits and tastes can be recovered just considering the bipartite network of users and products, in which links have different weights due to the score assigned to items. With respect to other weighted bipartite networks, in these systems we observe a maximum possible weight per link, that limits the variability of the outcomes. In the present article we propose an entropy-based randomisation of (bipartite) rating networks by extending the Configuration Model framework: the randomised network satisfies the constraints of the degree per rating, i.e. the number of given ratings received by the specified product or assigned by the single user. We first show that such a null model is able to reproduce several non-trivial features of the real network better than other null models. Then, using it as a benchmark, we project the information contained in the real system on one of the layers, showing, for instance, the division in communities of music albums due to the taste of customers, or, in movies due the audience.