GraphRec: Efficient and Scalable Recursive Programming with Graphs – GraphRec

Large graph data structures are increasingly used in many areas such as transportation networks, healthcare (drug interactions), cryptocurrencies (blockchain transactions), social networks, knowledge management, etc. Recursive programs constitute a very powerful means to extract valuable information from these linked data structures. However, recursive programs can be very costly to evaluate on large graphs and sometimes hardly feasible – if feasible at all – on modern graphs with millions of nodes and edges. This project is an endeavor to solve foundational and algorithmic challenges for optimizing expressive recursive programs, enabling new, more robust and more efficient value extraction methods from large graphs. We will investigate effective programming models, compilation techniques and static analyses for producing specialized language runtimes. We will study how to synthesize code which is correct and optimized for execution on distributed platforms. The overall expected outcome is to make the development of large-scale-graph-data-intensive applications less error-prone and more efficient.