Abstract
Perhaps the most widely appreciated linked data principle instructs linked data providers to provide useful information using the standards (i.e., RDF and SPARQL). Such information corresponds to patterns expressed as SPARQL queries that are matched against the RDF graph. Until recently, patterns had to specify the exact path that would match against the underlying graph. The advent of the SPARQL 1.1 Recommendation introduced property paths as a new graph matching paradigm that allows the employment of Kleene star * (and its variant Kleene plus +) unary operators to build SPARQL queries that are agnostic of the underlying RDF graph structure. In this paper, we present the Top-k Shortest Paths in large typed RDF Datasets Challenge. It highlights the key aspects of property path queries that employ the Kleene star operator, presenting three widely different approaches.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Zhang, X., Van den Bussche, J.: On the power of SPARQL in expressing navigational queries. Comput. J. 58(11), 2841–2851 (2014)
Gubichev, A., Bedathur, S.J., Seufert, S.: Sparqling kleene: fast property paths in RDF-3X. In: First International Workshop on Graph Data Management Experiences and Systems, GRADES 2013, pp. 14:1–14:7. ACM, New York (2013)
Arenas, M., Conca, S., Perez, J.: Counting beyond a yottabyte, or how SPARQL 1.1 property paths will prevent adoption of the standard. In: Proceedings of the 21st International Conference on World Wide Web, WWW 2012, pp. 629–638. ACM, New York (2012)
Arenas, M., Gutierrez, C., Miranker, D.P., Perez, J., Sequeda, J.F.: Querying semantic data on the web? ACM SIGMOD Rec. 41(4), 6–17 (2013)
Hassan, Z., Qadir, M. A., Islam, M. A., Shahzad, U., Akhter, N.: Modified MinG Algorithm to Find Top-K Shortest Paths from Large RDF Graphs. In: ESWC 2016
Hertling, S., Schroeder, M., Jilek, C., Dengel, A.: Top-K Shortest Paths in Directed, Labeled Multigraphs. In: ESWC 2016
De Vocht, L., Verborgh, R., Mannens, E, Van de Walle, R.: Using Triple Pattern Fragments to Enable Streaming of Top-K Shortest Paths via the Web. In: ESWC 2016
Barton, S.: Indexing graph structured data. Ph.D. thesis, Masaryk University, Brno, Czech Republic (2007)
Eppstein, D.: Finding the k shortest paths. SIAM J. Comput. 28(2), 652–673 (1998)
De Vocht, L., Beecks, C., Verborgh, R., Seidl, T., Mannens, E., Van de Walle, R.: Improving semantic relatedness in paths for storytelling with linked data on the web. In: Gandon, F., Guéret, C., Villata, S., Breslin, J., Faron-Zucker, C., Zimmermann, A. (eds.) ESWC 2015. LNCS, vol. 9341, pp. 31–35. Springer, Heidelberg (2015). doi:10.1007/978-3-319-25639-9_6
Salvadores, M., Horridge, M., Alexander, P.R., Fergerson, R.W., Musen, M.A., Noy, N.F.: Using SPARQL to Query BioPortal Ontologies and Metadata
Tarjan, R.E.: Fast algorithms for solving path problems. J. ACM 28(3), 594–614 (1981)
Dietz, P.F.: Maintaining order in a linked list. In: STOC 1982: Proceedings of the Fourteenth Annual ACM Symposium on Theory of Computing, pp. 122–127. ACM Press, New York (1982)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Annex
Annex
Practical example.
The image below (Fig. 2) depicts the example RDF dataset D1. Each node is a subject or object while each edge is a predicate.
Tiny example dataset
D1 contains exactly 4 paths from A to B:
-
9.
(A,P,u3,p7,B)
-
10.
(A,p3,u6,P,B)
-
11.
(A,p1,u1,p2,u2,p8,B)
-
12.
(A,P,u3,p4,u4,p5,u5,p6,u3,p7,B)
At this point, it should be mentioned that a path is valid only if it contains unique triples. For example, the path:
-
(A,P,u3,p4,u4,p5,u5,p6,u3,p4,u4,p5,u5,p6,u3,p7,B)
is not valid, since the triple: u3,p4,u4 exists more than once.
Task 1.
If task1 requires 3 paths between A and B within D1, the expected results should be the top-3 shortest paths from A to B:
-
1.
(A,P,u3,p7,B)
-
2.
(A,p3,u6,P,B)
-
3.
(A,p1,u1,p2,u2,p8,B)
Note that the first two paths have the same length (i.e., 2), since they both contain two edges). The third path has a length that equals to 3, so it comes after the first two paths.
Task 2.
If Task2 requires 2 paths between A and B within D1, the expected results should be the top-2 shortest paths from A to B that have P as their first or last edge:
-
4.
(A,P,u3,p7,B)
-
5.
(A,p3,u6,P,B)
If Task2 requires 3 paths between A and B within D1, the expected results should be the top-3 shortest paths from A to B that have P as their first or last edge:
-
6.
(A,P,u3,p7,B)
-
7.
(A,p3,u6,P,B)
-
8.
(A,P,u3,p4,u4,p5,u5,p6,u3,p7,B)
Note that the path (A,p1,u1,p2,u2,p8,B) is omitted since P is neither the first nor the last edge of the path, so it does not fit into the requirements of the path pattern.
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this paper
Cite this paper
Papadakis, I., Stefanidakis, M., Mylonas, P., Niggemeyer, B.E., Kazanas, S. (2016). Top-K Shortest Paths in Large Typed RDF Datasets Challenge. In: Sack, H., Dietze, S., Tordai, A., Lange, C. (eds) Semantic Web Challenges. SemWebEval 2016. Communications in Computer and Information Science, vol 641. Springer, Cham. https://doi.org/10.1007/978-3-319-46565-4_15
Download citation
DOI: https://doi.org/10.1007/978-3-319-46565-4_15
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-46564-7
Online ISBN: 978-3-319-46565-4
eBook Packages: Computer ScienceComputer Science (R0)