Compression and Reordering for Parallel Graph Analytics
We plan to develop a suite of graph compression and reordering techniques as part of the Ligra parallel graph processing framework to reduce space usage and improve performance of graph algorithms.
Graphs algorithms have many applications, such as in analyzing social networks, biological networks, and unstructured meshes in scientific simulations. Due to the recent growth in data sizes, improving the running time and space usage of graph algorithms has become very important. This project aims to research graph compression and reordering techniques to reduce the space usage and improve the running time of parallel graph algorithms. These techniques will be integrated into the Ligra graph processing framework to enable users to easily use them in a large class of graph algorithms.