Tink, a temporal graph analytics library for Apache Flink

Back to Sessions overview

The study of temporal graphs is still in its infancy. Hence not many libraries or tools have been developed to support this type of analysis. Most existing tools for temporal graph analysis are built standalone, whereas Tink, a library for temporal graph analytics, is built on top of Apache Flink and, therefore, leverages its advanced features such as distributed processing and query optimization plans. A temporal graph extends a regular graph, therefore Tink uses most of Gelly's structure when reasoning about graphs. Tink focuses on interval graphs in which every edge is associated with a starting time and an ending time. This subsumes most current research which focuses on sequence graphs in which edges have only a single starting time. Tink enables a variety of temporal graph algorithms, it is built as a framework and comes with several example algorithms such as temporal shortest paths and different temporal centrality metrics. Tink is open source and can be used free of charge