Home  /  Ricerca  /  Fuzzy clustering of spatial interval-valued data

Fuzzy clustering of spatial interval-valued data

In this article, developed in collaboration with the Department of Social and Economic Sciences at La Sapienza University, we create a classification algorithm for units characterized using the maximum and minimum of specific attributes along with a network of adjacency relations. As part of the LATIF project and the MediaFutures project, we test this algorithm on data related to the Twitter activity of accounts recognized as members of the International Fact-Checking Network. What we achieve is the division of accounts into 3 groups: accounts with a common audience, accounts with each having their isolated audience, and accounts with minimal resonance.

Read the full paper HERE

By Pierpaolo D’Urso, Livia De Giovanni, Lorenzo Federico, and Vincenzina Vitale