End-to-end Graph Analytics

Full Name: Manos Papagelis

Academic Affiliation: York University

Position: Assistant Professor

Abstract: Increasingly, organizations and communities are turning to big data analytics and network analysis to make sense of their data, solve computational problems and inform faster and better decisions that might have a business and/or societal impact. Graph mining and exploration, built on the mathematics of graph theory, leverage network structures to model and analyze pairwise relationships between objects (or people). With a growing number of networks – social, technological, biological – becoming available and representing an ever increasing amount of information, the ability to easily perform large-scale graph analytics is key to revealing the underlying dynamics of these networks, not easily observable before. Current practice requires technical expertise and domain-knowledge in order to carry out the various steps of data ingestion, modeling and processing that is involved, before (multiple iterations of) analysis can take place. This renders the whole process of graph mining and exploration cumbersome for non-experts and limits its applicability. The long-term objective of our research is the development of a novel, unifying framework for performing end-to-end graph analytics. That involves developing the theory and tools for making the graph mining and exploration process simple and flexible, so it can be easily applied to diverse problem settings and domains. The scientific approach of our research is characterized by acquisition, processing, modeling, analysis and visualization of very large graph data sets and involves the use of advanced techniques for data analysis, such as text analytics, machine learning, predictive analytics, and natural language processing.