Building Decision-making Indicators Through Network Analysis of Big Data

In recent years we have witnessed a growing concern in scientific research to understand and improve actionable analytics-driven decision processes, mostly focused on online data. Many researchers have focused their attention on computational and Information and Communications Technology issues in this matter. Only a small share of literature is concerned with how indicators can be improved by Big Data analytics. In this paper, we propose an innovative methodological approach to building indicators by combining Big Data analytics with the analysis of network patterns. Our study aims to define relational structures in a Big Data set, implementing measurements and clustering methods by Network Analysis in order to build decision-making indicators. We describe an audience model both to collect a large amount of online data from large online newspapers and to structure those in a relational form. By analysing readers' comments, we can derive proxies of reliable indicators about specific topics discussed on an online newspaper blog. We show the effectiveness of such an approach in detecting and building indicators to support policy-makers in complex decision-making processes.