An evaluation of approaches to federated query processing over linked data

The W3C SPARQL working group is defining the new SPARQL 1.1 query language. The current working draft of SPARQL 1.1 focuses mainly on the description of the language. In this paper, we provide a formalization of the syntax and semantics of the SPARQL 1.1 federation extension, an important fragment of the language that has not yet received much attention. Besides, we propose optimization techniques for this fragment, provide an implementation of the fragment including these techniques, and carry out a series of experiments that show that our optimization procedures could significantly speed up the query evaluation process.

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Semantics and Optimization of the SPARQL 1.1 Federation Extension

Buil-Aranda

Arenas²,

Corcho

2011

“…For many queries proposed in [6] a performance gain of more than 90% can be achieved resulting in improvements of an order of magnitude, timeouts do not occur any longer. This is in particular due to the improved join order and the other above mentioned optimizations.…”

Section: Fedx -Design and System Overviewmentioning

confidence: 99%

“…In practical terms the Sesame framework in conjunction with AliBaba 1 is one possible sample solution allowing for federations of distributed repositories and endpoints. However, benchmarks have shown poor performance for many queries in the federated setup due to the absence of advanced optimization techniques [6]. From the research community DARQ [9] and Networked Graphs [10] contribute approaches to federated SPARQL queries and federated integration.…”

Section: Introductionmentioning

confidence: 99%

FedX: A Federation Layer for Distributed Query Processing on Linked Open Data

Schwarte

Haase

Hose

et al. 2011

Self Cite

Abstract. Driven by the success of the Linked Open Data initiative today's Semantic Web is best characterized as a Web of interlinked datasets. Hand in hand with this structure new challenges to query processing are arising. Especially queries for which more than one data source can contribute results require advanced optimization and evaluation approaches, the major challenge lying in the nature of distribution: Heterogenous data sources have to be integrated into a federation to globally appear as a single repository. On the query level, though, techniques have to be developed to meet the requirements of efficient query computation in the distributed setting. We present FedX, a project which extends the Sesame Framework with a federation layer that enables efficient query processing on distributed Linked Open Data sources. We discuss key insights to its architecture and summarize our optimization techniques for the federated setting. The practicability of our system will be demonstrated in various scenarios using the Information Workbench.

“…To the best of our knowledge, the only work addressing the latter issue was our previous work in [13], which -focusing on selected federation approachesserved as a starting point for designing FedBench. Going far beyond this initial work, in this paper we present a holistic benchmark suite, including a variety of new data and query sets, new scenarios such as Linked Data access (i.e., via HTTP requests), an automated evaluation framework (with support for various metrics like counting the number of requests, automated evaluation, interfaces for connecting new systems etc.…”

Section: Introductionmentioning

confidence: 99%

FedBench: A Benchmark Suite for Federated Semantic Data Query Processing

Schmidt

Görlitz

Haase

et al. 2011

Self Cite

117

108

Abstract. In this paper we present FedBench, a comprehensive benchmark suite for testing and analyzing the performance of federated query processing strategies on semantic data. The major challenge lies in the heterogeneity of semantic data use cases, where applications may face different settings at both the data and query level, such as varying data access interfaces, incomplete knowledge about data sources, availability of different statistics, and varying degrees of query expressiveness. Accounting for this heterogeneity, we present a highly flexible benchmark suite, which can be customized to accommodate a variety of use cases and compare competing approaches. We discuss design decisions, highlight the flexibility in customization, and elaborate on the choice of data and query sets. The practicability of our benchmark is demonstrated by a rigorous evaluation of various application scenarios, where we indicate both the benefits as well as limitations of the state-of-the-art federated query processing strategies for semantic data.