Aggregation and Disaggregation of Information: A Holistic View

Abstract

Data-driven analytics and decision-making have been essential for numerous applications in our society. To transform the data into a source of rich intelligence and support decision-making, data-driven analytics often need to aggregate intelligence from multiple sources and disaggregate signals into significant constituents. Though many existing approaches perform these two tasks respectively, there are few attempts to study them with a holistic view. This dissertation exploits the intrinsic connections between intelligence aggregation and signal disaggregation by developing novel models to capture and leverage various types of non-IIDness and inter-correlations in the data from complex systems. Our preliminary results show that, by identifying non-IIDness of information sources, our approach outperforms alternative methods for intelligence aggregation tasks. Also, by viewing disaggregation as the inverse function of aggregation and incorporating various types of inter-correlations in complex systems, we can also improve the performance for signal disaggregation tasks. Given these promising results, we will further improve the effectiveness and efficiency of our framework on large-scale data from different application fields.