Temporal property graphs are graphs whose structure and properties change over time. Temporal graph datasets tend to be large due to stored historical information, asking for scalable analysis capabilities. In this talk, Christopher Rost gives an overview of Gradoop, a graph dataflow system for scalable, distributed analytics of temporal property graphs which has been continuously developed since 2015 at the University of Leipzig. He will present the system architecture of Gradoop, its data model TPGM with composable temporal graph operators and several implementation details. The talk ends with a short performance evaluation and a reflection on lessons learned from the Gradoop effort.
Bio: Christopher Rost, born 1990 in Germany Master degree computer science at University of Leipzig Since 2018 PhD student at University of Leipzig Research interests: Distributed graph processing, Temporal graph maintenance, querying, and analysis, Graph stream processing and analysis Teaching: Big Data practical course, Data Science workshop Recent publication - Distributed temporal graph analytics with GRADOOP - VLDB Journal 2021 Special Issue Paper
