Description
The emergence of services and applications, such as social networks, World Wide Web, semantic web with knowledge graphs and RDF, genome and scientific databases, etc., projects to the frontend graph structured data as a precious tool to organise data but with new perspectives and challenges mainly related to handling massive data. In fact, the volume of data increases to such an extent that even polynomial solutions are no longer sufficient. Distributed and parallel frameworks, which are effective approaches to deal with big data, are not necessarily suitable for big graph computations mainly because of the structure of graph datasets and the iterative nature of their algorithms.
In this context, graph compression provides another point of view from which we can deal with this issue. It consists to produce small and simple representations of graphs on which we can undertake complex graph operations and verify or estimate complex graph properties.
Several research domains and applications rely on the computation of such summaries and compact representations: graph learning, graph indexing and querying, graph analysis and visualization, etc. This explains why graph summarizing is gaining importance. However, this task is complex and relies on a deep knowledge of graphs, their properties and their applications.
The objective of the Coregraphie project is to construct graph summaries and design the algorithms that allow to query these summaries.
Coregraphie is a four-year project funded by the French National Agency for Research (ANR), starting April 2021.
