• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ³í¹®Áö D : µ¥ÀÌŸº£À̽º

Á¤º¸°úÇÐȸ ³í¹®Áö D : µ¥ÀÌŸº£À̽º

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) º¹ÇÕ À̺¥Æ® Ã³¸®¸¦ À§ÇÑ °í¼Ó ¹úÅ©·Îµù Áö¿ø ´ë¿ë·® Æ®¸®ÇàÀúÀå¼Ò
¿µ¹®Á¦¸ñ(English Title) Large Triple Store Supporting Fast Bulk-Loading for Complex Event Processing
ÀúÀÚ(Author) ¾öÁ¤È£   ±èÅÂÈ«   Ȳ¹Ì³ç   ¼­µ¿¹Î   ÃÖ¼ºÇÊ   À̽¿젠 Á¤ÇѹΠ  Jung-Ho Um   Taehong Kim   Mi-Nyeong Hwang   Dongmin Seo   Sungpil Choi   Seungwoo Lee   Hanmin Jung  
¿ø¹®¼ö·Ïó(Citation) VOL 41 NO. 03 PP. 0174 ~ 0180 (2014. 06)
Çѱ۳»¿ë
(Korean Abstract)
IT ¹× °úÇÐ ±â¼úÀÇ ¹ßÀüÀ¸·Î ¼ö¸¹Àº µ¥ÀÌÅÍ°¡ »ý¼ºµÇ´Â ºòµ¥ÀÌÅÍ ½Ã´ë°¡ µµ·¡ÇÔ¿¡ µû¶ó ºòµ¥ÀÌÅ͸¦ ÀÌ¿ëÇÏ¿© ºÐ¼®ÇÏ´Â ±â¼úÀÌ ÁÖ¸ñÀ» ¹Þ°í ÀÖ´Ù. ÀÌ Áß¿¡¼­ Áö½ÃÀû ºÐ¼®(Prescriptive Analytics) ±â¼úÀº ÇöÀçÀÇ ÇöȲÀ» ºÐ¼®ÇÏ¿© »ç¿ëÀÚÀÇ ¹Ì·¡¿¡ ´ëÇÏ¿© ¹æÇâÀ» Á¦½ÃÇÏ´Â Â÷¼¼´ë ºÐ¼® ±â¼úÀÌ´Ù. À̸¦ À§Çؼ­´Â ´Ù¾çÇÑ µ¥ÀÌÅÍ ¿øõÀ¸·ÎºÎÅÍÀÇ ºÐ¼®À» Åä´ë·Î º¹ÇÕ À̺¥Æ® 󸮰¡ ÇÊ¿äÇÏ´Ù. ÀÌ´Â ´Ù¾çÇÑ µ¥ÀÌÅÍ ¿øõÀ¸·ÎºÎÅÍ ÃßÃ⠶Ǵ »ý¼ºµÈ Æ®¸®Çà µ¥ÀÌÅÍÀÇ ºÐ¼® Á¤º¸¸¦ ±â¹ÝÀ¸·Î ÇÑ´Ù. º» ³í¹®¿¡¼­´Â ÀÌ·¯ÇÑ ´ë¿ë·® Æ®¸®Çà µ¥ÀÌÅ͸¦ È¿À²ÀûÀ¸·Î ÀúÀåÇÏ´Â ±â¹ýÀ» Á¦¾ÈÇÏ°í, ¾Æ¿ï·¯ »ç¿ëÀÚ¿¡°Ô ºü¸£°Ô °Ë»öÀÌ °¡´ÉÇÑ ´ë¿ë·® Æ®¸®Çà ÀúÀå¼Ò¸¦ ¼³°èÇÑ´Ù. À̸¦ À§ÇØ Á¦¾ÈÇÏ´Â ÀúÀå¼Ò´Â Æ®¸®Çà ÀúÀåÀ» À§ÇØ ºÐ»ê µ¥ÀÌÅͺ£À̽º¸¦ »ç¿ëÇÏ°í, ¸Ê¸®µà½º¸¦ È°¿ëÇÑ ¹úÅ©·Îµù ¾Ë°í¸®ÁòÀ» ¼³°èÇÏ¿´À¸¸ç, Æ®¸®Çà °Ë»öÀ» À§ÇØ SPARQL ÁúÀÇ Ã³¸® ¿£Áø°ú ºÐ»ê µ¥ÀÌÅͺ£À̽º¸¦ ¿¬°è ¹× °³¹ßÇÏ¿´´Ù. ¼º´ÉÆò°¡¸¦ ÅëÇØ Á¦¾ÈÇÏ´Â ¹úÅ© ·Îµù ±â¹ýÀÌ ÀúÀå ¼Óµµ¿¡¼­´Â ±âÁ¸ ½Ã½ºÅÛ¿¡ ºñÇØ 2.5¹èÀÇ ¼º´É Çâ»óÀ» º¸¿´À¸¸ç, ´Ù¾çÇÑ ´ë¿ë·®ÀÇ Æ®¸®Çà µ¥ÀÌÅÍ ¼Â¿¡ ´ëÇؼ­µµ È¿À²ÀûÀ¸·Î ÀúÀåÀÌ °¡´ÉÇÔÀ» º¸¿´´Ù. ¾Æ¿ï·¯, Àüü µ¥ÀÌÅÍ ¼Â¿¡ ´ëÇÏ¿© select ÁúÀÇ ½Ã °Ë»ö ¼º´ÉÀº Æò±Õ 1ÃÊ·Î º¹ÇÕ À̺¥Æ® 󸮸¦ À§ÇÑ °í¼Ó °Ë»öÀÌ °¡´ÉÇÔÀ» ¾Ë ¼ö ÀÖ¾ú´Ù.
¿µ¹®³»¿ë
(English Abstract)
Due to the development of IT and scientific technology, huge amounts of data are created so that big data age has come. Among many big data techniques, big data analytics technique is concentrated on. Especially, prescriptive analytics techniques are next analytics technique because it can give the future's directions by analyzing current status. For this, it needs complex event processing which analyzes the data from many data sources. Complex event processing is based on the analyzing information from triple data or extraction data on a variety of data sources. In this paper, we propose the bulk-loading technique for large triple enabling to store the data efficiently and design triple store to respond the user's query quickly. For this, the proposed triple store uses distributed database. And we design bulk-loading algorithm using MapReduce framework and develop SPARQL query processing engine connecting to the distributed database. Experimental results show that the proposed bulk-loading technique has 2.5 times better than legacy system in terms of loading performance. Also, the results show that the bulk-loading technique can store various large triple data set efficiently. The select query response for whole data set is approximately 1 second. It shows that the proposed triple store supports fast search for complex event processing.
Å°¿öµå(Keyword) Æ®¸®Çà ÀúÀå¼Ò   MapReduce   Hbase   LOD   º¹ÇÕ À̺¥Æ® 󸮠  TripleStore   MapReduce   Hbase   LOD   complex event processing  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå