Functional Dependencies Unleashed for Scalable Data Exchange

Abstract

International audienceWe address the problem of efficiently evaluating target functional dependencies (fds) in the Data Exchange (DE) process. Target fds naturally occur in many DE scenarios, including the ones in Life Sciences in which multiple source relations need to be structured under a constrained target schema. However, despite their wide use, target fds' evaluation is still a bottleneck in the state-of-the-art DE engines. Systems relying on an all-SQL approach typically do not support target fds unless additional information is provided. Alternatively, DE engines that do include these dependencies typically pay the price of a significant drop in performance and scalability. In this paper, we present a novel chase-based algorithm that can efficiently handle arbitrary fds on the target. Our approach essentially relies on exploiting the interactions between source-to-target (s-t) tuple-generating dependencies (tgds) and target fds. This allows us to tame the size of the intermediate chase results, by playing on a careful ordering of chase steps interleaving fds and (chosen) tgds. As a direct consequence, we importantly diminish the fd application scope, often a central cause of the dramatic overhead induced by target fds. Moreover, reasoning on dependency interaction further leads us to interesting parallelization opportunities, yielding additional scalability gains. We provide a proof-of-concept implementation of our chase-based algorithm and an experimental study aimed at gauging its scalability and efficiency. Finally, we empirically compare with the latest DE engines, and show that our algorithm outperforms them

Benchmarking the Chase

“…SELECT DISTINCT R.a, append(' _ Sk _ f(',R.a,')') FROM R (10) The parallel chase [11] weakens line 6 to check whether h is active in I, rather than in µ(I ∪ N); since I is fixed in an iteration, this can make checking active triggers much easier to implement. Known acyclicity conditions [18] ensure termination of the parallel chase, and the solution is deterministic, although it may be larger than the one produced by the restricted chase.…”

“…CHASEFUN [10] is a more recent data exchange system. It supports only s-t TGDs and functional dependencies, and it implements a variant of the unrestricted Skolem chase in which TGD and EGD chase steps are ordered to reduce the size of the intermediate chase results.…”

“…Experimental evaluations of many data exchange systems have already been conducted [34,28,27,15,20,10]. In many cases we reused and extended systems from these earlier studies, but our work differs in several important ways.…”

Parallel Reasoning of Graph Functional Dependencies

“…Related also the chase [29], a classical tool in the relational dependency theory. The chase has been employed in data exchange [30], [31], data repairing [32] and query rewriting [33], with relational tuple-generating dependencies (TGDs) and equality generating dependencies (EGD; see [34] for a survey). As remarked earlier, the chase has also been studied for FDs on RDF [4], [5] and for GEDs [2].…”

JOWSDATALOG: An Ontology Based Data Access Tool for Jointly-Weakly-Sticky Datalog± Programs

“…We want to review OBDA tools that can deal with Datalog ± programs, and their chase procedure is integrated with an RDBMS. The interesting research that we found is [Benedikt et al, 2017] which provides a benchmark, ChaseBench, by running a variety range of experiments over different tools, ChaseFun [Bonifati et al, 2016], DEMo [Pichler and Savenkov, 2009], Llunatic [Geerts et al, 2013], and PDQ [Benedikt et al, 2015]. The primary reason that some of these tools are proposed is solving the problem of data exchange (DE).…”