An Empirical Evaluation of RDF Graph Partitioning Techniques

Akhter, Adnan; Ngonga, Axel-Cyrille Ngomo; Saleem, Muhammad

doi:10.1007/978-3-030-03667-6_1

Cited by 23 publications

(31 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, create a separate partition of all RDF triples with Predicate age and object values between 30 and 40. In our motivating example, let the partition key is the triple number with partitions defined according the following ranges: first partition is created for all the triples in the range [1,4], a second partition is created for all the triples in the range [5,8], and third partition is created for all the triples in the range [9,11]. Workload-Based Partitioning: The partitioning techniques in this category make use of the query workload to partition the given RDF dataset.…”

Section: Rdf Graph Partitioningmentioning

confidence: 99%

“…This technique is exhibited in example given in Figure 12, shows all the triples having hierarchy1 in subjects are assigned to the green partition, triples having hierarchy2 in subjects are assigned to the red partition, and triples having hierarchy3 in subjects are assigned to the blue partition. This partitioning may not produce the best query runtimes as the underlying assumptions about IRIs might not be true in practice [4].…”

Section: Rdf Graph Partitioningmentioning

confidence: 99%

“…Data storage is dependent on factors like the format of storage, size of the storage and inference supported by the storage format [81]. A recent evaluation [4] shows that the storage of RDF data in different RDF graph partitioning techniques has a vital effect on the query runtime.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Storage, Indexing, Query Processing, and Benchmarking in Centralized and Distributed RDF Engines: A Survey

Ali¹,

Saleem²,

Yao³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

The recent advancements of the Semantic Web and Linked Data have changed the working of the traditional web. There is significant adoption of the Resource Description Framework (RDF) format for saving of web-based data. This massive adoption has paved the way for the development of various centralized and distributed RDF processing engines. These engines employ various mechanisms to implement critical components of the query processing engines such as data storage, indexing, language support, and query execution. All these components govern how queries are executed and can have a substantial effect on the query runtime. For example, the storage of RDF data in various ways significantly affects the data storage space required and the query runtime performance. The type of indexing approach used in RDF engines is critical for fast data lookup. The type of the underlying querying language (e.g., SPARQL or SQL) used for query execution is a crucial optimization component of the RDF storage solutions. Finally, query execution involving different join orders significantly affects the query response time. This paper provides a comprehensive review of centralized and distributed RDF engines in terms of storage, indexing, language support, and query execution.

show abstract

Section: Rdf Graph Partitioningmentioning

confidence: 99%

Section: Rdf Graph Partitioningmentioning

confidence: 99%

See 1 more Smart Citation

Storage, Indexing, Query Processing, and Benchmarking in Centralized and Distributed RDF Engines: A Survey

Ali¹,

Saleem²,

Yao³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Thus by using the first come for serve strategy, all the triples with predicate p1 will be assigned to first partition (red), p2 triples will be assigned to second partition (green), p3 triples will be assigned to third partition (blue), and p4 triples will be again assigned to first partition. This technique can leads to significant performance improvement, provided that the predicates are intelligently grouped intro partitions, such that communication load among data nodes is reduced [11].…”

Section: Fig 11 Types Of Partitioning Used Rdf Enginesmentioning

confidence: 99%

Storage, Indexing, Query Processing, and Benchmarking in Centralized and Distributed RDF Engines: A Survey

Ali¹,

Yao²,

Ngomo³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

The recent advancements of the Semantic Web and Linked Data have changed the working of the traditional web. There is a huge adoption of the Resource Description Framework (RDF) format for saving of web-based data. This massive adoption has paved the way for the development of various centralized and distributed RDF processing engines. These engines employ different mechanisms to implement key components of the query processing engines such as data storage, indexing, language support, and query execution. All these components govern how queries are executed and can have a substantial effect on the query runtime. For example, the storage of RDF data in various ways significantly affects the data storage space required and the query runtime performance. The type of indexing approach used in RDF engines is key for fast data lookup. The type of the underlying querying language (e.g., SPARQL or SQL) used for query execution is a key optimization component of the RDF storage solutions. Finally, query execution involving different join orders significantly affects the query response time. This paper provides a comprehensive review of centralized and distributed RDF engines in terms of storage, indexing, language support, and query execution.

show abstract

“…Akhter et al compared seven RDF graph partitioning methods in two different evaluation setups [45]. The existing partitioning methods are evaluated in terms of partitioning time, partitioning imbalance, and query processing time.…”

mentioning

confidence: 99%

Dynamic Partitioning Supporting Load Balancing for Distributed RDF Graph Stores

2019

View full text Add to dashboard Cite

Various resource description framework (RDF) partitioning methods have been studied for the efficient distributed processing of a large RDF graph. The RDF graph has symmetrical characteristics because subject and object can be used interchangeably if predicate is changed. This paper proposes a dynamic partitioning method of RDF graphs to support load balancing in distributed environments where data insertion and change continue to occur. The proposed method generates clusters and subclusters by considering the usage frequency of the RDF graph that are used by queries as the criteria to perform graph partitioning. It creates a cluster by grouping RDF subgraphs with higher usage frequency while creating a subcluster with lower usage frequency. These clusters and subclusters conduct load balancing by using the mean frequency of queries for the distributed server and conduct graph data partitioning by considering the size of the data stored in each distributed server. It also minimizes the number of edge-cuts connected to clusters and subclusters to minimize communication costs between servers. This solves the problem of data concentration to specific servers due to ongoing data changes and additions and allows efficient load balancing among servers. The performance results show that the proposed method significantly outperforms the existing partitioning methods in terms of query performance time in a distributed server. This paper is organized as follows. Section 2 describes the existing graph partitioning methods. Section 3 describes the RDF dynamic partitioning method proposed in this paper. Section 4 describes the results of the performance evaluation to show the excellence of the proposed method. Finally, Section 5 provides the conclusions of this paper and presents future directions. Related WorkHendrickson and Leland proposed a multilevel method for graph partitioning within acceptable execution time [23]. The multilevel partitioning method generates a sequence of smaller graphs approximating the original graph to find easily a good partition via coarsening procedure. In the coarsening step, it merges two vertices connected to an edge and the new vertex retains edges to the union of the neighbors of the merged vertices. The smallest graph is partitioned via a spectral partitioner. The spectral partitioner needs one, two, or three eigenvectors of the Laplacian matrix to partition a graph into two, four, or eight groups, respectively. The partitioning algorithm is able to handle edge and vertex weights, even if the original graph is unweighted. The multilevel partitioning method makes the sums of the vertex weights in each server as equal as possible and the sum of the weights of edges crossing between servers is minimized.Wang and Chiu proposed a RDF graph partitioning method to minimize the number of edges on different servers [40]. This method first scans the RDF graph and stores subgraph with common vertices into the same components. If the connected components of the graph are found, each connected componen...

show abstract

An Empirical Evaluation of RDF Graph Partitioning Techniques

Cited by 23 publications

References 20 publications

Storage, Indexing, Query Processing, and Benchmarking in Centralized and Distributed RDF Engines: A Survey

Storage, Indexing, Query Processing, and Benchmarking in Centralized and Distributed RDF Engines: A Survey

Storage, Indexing, Query Processing, and Benchmarking in Centralized and Distributed RDF Engines: A Survey

Dynamic Partitioning Supporting Load Balancing for Distributed RDF Graph Stores

Contact Info

Product

Resources

About