Topology reduction and probabilistic information extraction for large data-sets: A disaster management case study

The dynamical and probabilistic properties of the relationships among data modelled by real-world networks have drawn extensive research from a several interdisciplinary fields. They, in fact, can successfully identify the main properties of large data-sets. However, a deep analysis of such networks is likely to generate information of little use due to their inherent complexity, as well as the inconsistencies of data modelled by them. In this paper, we discuss the evaluation of a method as part of ongoing research which aims to extract, assess and identify relevant information based on the mutual probabilistic relationships among the data captured by Big Data. In order to validate and support our approach, a large dataset capturing information on the seismological activity provided by the European-Mediterranean Seismological Centre is considered. We will show that this approach provides a scalable, accurate and useful tool to enhance the state of the art research within disaster management. The approach discussed in this paper further supports our effort to create a big data analytics tool aiming to extract actionable intelligence from a variety of big datasets.