Title:Efficient RDF Streaming for the Edge-Cloud Continuum

Abstract:With the ongoing, gradual shift of large-scale distributed systems towards the edge-cloud continuum, the need arises for software solutions that are universal, scalable, practical, and grounded in well-established technologies. Simultaneously, semantic technologies, especially in the streaming context, are becoming increasingly important for enabling interoperability in edge-cloud systems. However, in recent years, the field of semantic data streaming has been stagnant, and there are no available solutions that would fit those requirements. To fill this gap, in this contribution, a novel end-to-end RDF streaming approach is proposed (named Jelly). The method is simple to implement, yet very elastic, and designed to fit a wide variety of use cases. Its practical performance is evaluated in a series of experiments, including end-to-end throughput and latency measurements. It is shown that Jelly achieves vastly superior performance to the currently available approaches. The presented method makes significant progress towards enabling high-performance semantic data processing in a wide variety of applications, including future edge-cloud systems. Moreover, this study opens up the possibility of applying and evaluating the method in real-life scenarios, which will be the focus of further research.