download0 view212
twitter facebook

공공누리This item is licensed Korea Open Government License

dc.contributor.author
엄정호
dc.contributor.author
정창후
dc.contributor.author
김태홍
dc.contributor.author
송사광
dc.contributor.author
이승우
dc.contributor.author
정한민
dc.date.accessioned
2019-08-28T07:41:53Z
dc.date.available
2019-08-28T07:41:53Z
dc.date.issued
2016-02-16
dc.identifier.issn
0920-8542
dc.identifier.uri
https://repository.kisti.re.kr/handle/10580/14514
dc.identifier.uri
http://www.ndsl.kr/ndsl/search/detail/article/articleSearchResultDetail.do?cn=NART75711639
dc.description.abstract
As the development of IT and scientific technology, very large amountsof knowledge data are continuously being created and the big data era can be saidto have arrived. Therefore, RDF store inserting and inquiring into knowledge baseshas to be scaled up in order to deal with such large sources of data. To this end, wepropose a scalable distributed RDF store based on a distributed database that usesbulk-loading for billions of triples to store data and to respond to user queries quickly.In order to achieve this purpose, we introduce a bulk-loading algorithm using theMapReduce framework and the SPARQL query processing engine to connect to alarge distributed database. Experimental results show that the proposed bulk-loadingalgorithm achieves 67.893K triples per second to load approximately 33 billion triples.Therefore, the experiment proves proposed RDF store can manage billions of triplesscale data.
dc.language
eng
dc.relation.ispartofseries
Journal of Supercomputing
dc.title
Distributed RDF store for efficient searching billions of triples based on Hadoop
dc.citation.endPage
1840
dc.citation.number
5
dc.citation.startPage
1825
dc.citation.volume
72
dc.subject.keyword
Triple Store
dc.subject.keyword
Hbase
dc.subject.keyword
MapReduce
dc.subject.keyword
RDF
dc.subject.keyword
LOD
Appears in Collections:
7. KISTI 연구성과 > 학술지 발표논문
Files in This Item:
There are no files associated with this item.

Browse