µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
ºÐ»ê ȯ°æ¿¡¼ ´ë¿ë·® RDF µ¥ÀÌÅÍÀÇ »ç¿ë ÆÐÅÏÀ» °í·ÁÇÑ µ¿Àû ºÐÇÒ ±â¹ý |
¿µ¹®Á¦¸ñ(English Title) |
Dynamic Partitioning Scheme Considering Usage Patterns of Large-Scale RDF Data in Distributed Environments |
ÀúÀÚ(Author) |
±è¹Î¼ö
¼ÛÁø¿ì
ÀüÇö¿í
Á¤ÀçÀ±
ÀÓÁ¾ÅÂ
º¹°æ¼ö
À¯Àç¼ö
Minsoo Kim
JinWoo Song
Hyeonwook Jeon
Jaeyun Jeong
Jongtae Lim. Kyoungsoo Bok
Jaesoo Yoo
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 32 NO. 03 PP. 0035 ~ 0049 (2016. 12) |
Çѱ۳»¿ë (Korean Abstract) |
ºÐ»ê ȯ°æ¿¡¼ ƯÁ¤ ¼¹ö¿¡ ¹ß»ýÇÑ ºÎÇÏ ¶Ç´Â ¼¹ö °£ÀÇ Á¶ÀÎÀ¸·Î ¹ß»ýÇÑ ºÎÇϸ¦ ÇØ°áÇϱâ À§ÇÑ µ¿Àû ºÐÇÒÀÌ ÇÊ¿äÇÏ´Ù. º» ³í¹®¿¡¼´Â RDF µ¥ÀÌÅÍÀÇ »ç¿ë ÆÐÅÏÀ» °í·ÁÇÏ¿© ºÎÇÏ ºÐ»êÀ» ¼öÇàÇÏ´Â µ¿Àû ºÐÇÒ ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. Á¦¾ÈÇÏ´Â ±â¹ýÀº ±×·¡ÇÁ ºÐÇÒÀ» ¼öÇàÇϱâ À§ÇØ ºó¹øÇÑ ÁúÀÇÀÇ µ¥ÀÌÅ͸¦ ±â¹ÝÀ¸·Î Ŭ·¯½ºÅ͸¦ »ý¼ºÇÑ´Ù. ƯÁ¤ ¼¹ö¿¡ ¹ß»ýÇÑ ºÎÇϸ¦ ÁÙÀ̱â À§ÇØ Å¬·¯½ºÅÍ¿¡¼ ¿ÜºÎ ¼¹ö¿Í ¿¬°áµÈ °£¼± ¼ö°¡ ¸¹À¸¸é ÇØ´ç Ŭ·¯½ºÅ͸¦ ÀÌÁÖ½ÃÅ°°í ¿¬°áµÈ °£¼± ¼ö°¡ ÀûÀ¸¸é º¹Á¦ÇÑ´Ù. ¶ÇÇÑ ¼¹ö °£ÀÇ Á¶ÀÎÀ¸·Î ÀÎÇÑ Åë½Å·®À» ÁÙÀ̱â À§ÇØ ÇϳªÀÇ ¼¹ö·Î Ŭ·¯½ºÅ͸¦ ÀÌÁÖÇÏ¿© ºÐÇÒÀ» ¼öÇàÇÑ´Ù. Á¦¾ÈÇÏ´Â ±â¹ýÀÇ ¿ì¼ö¼ºÀ» ÀÔÁõÇϱâ À§ÇØ ÀÀ´ä ½Ã°£ ¹× ¼¹öÀÇ ºÎÇÏ ºñÀ²¿¡ ´ëÇÑ ¼º´É Æò°¡¸¦ ¼öÇàÇÑ´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
In distributed environments, dynamic partitioning is needed to solve a load on a particular server or a load caused by a join among servers. In this paper, we propose a dynamic partitioning scheme for load balancing by considering the usage patterns of RDF data. The proposed scheme generates clusters based on the data of frequent queries in order to perform graph partitioning. In order to reduce a load on a particular server, the proposed scheme moves or replicates a cluster when the number of edges in the cluster connected to external servers is over or under threshold. In addition, in order to reduce the amount of communication data by a join among servers, the proposed scheme moves the cluster to a server and partitions it. To show the superiority of the proposed scheme, we evaluate performance evaluations in terms of a response time and the load ratio of a server.
|
Å°¿öµå(Keyword) |
RDF
µ¿Àû ºÐÇÒ
ºÎÇÏ ºÐ»ê
Ŭ·¯½ºÅÍ
RDF
dynamic partitioning
load balancing
cluster
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|