Characteristic sets profile features: Estimation and application to SPARQL query planning

RDF dataset profiling is the task of extracting a formal representation dataset’s features. Such features may cover various aspects ranging from information on licensing and provenance to statistical descriptors data distribution its semantics. In this work, we focus characteristics sets profile that capture both structural semantic an dataset, making them valuable resource for different downstream applications. While previous research demonstrated benefits characteristic in centralized federated query processing, access these fine-grained statistics taken granted. However, especially computing feature challenging as it can be difficult and/or costly process entire all federation members. We address shortcoming by introducing concept estimation propose sampling-based approach generate estimations feature. addition, showcase applicability querying proposing planning specifically designed leverage estimations. our first experimental study, intrinsically evaluate representativeness estimation. The results show even small samples just 0.5 % original graph’s entities allow estimating properties Our second study extrinsically evaluates investigating their planner using well-known FedBench benchmark. experiments estimated obtaining efficient plans.