Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)
ÇѱÛÁ¦¸ñ(Korean Title) |
´ë¿ë·® µ¥ÀÌÅÍ ±â¹Ý SPARQL ÁúÀÇ°á°ú¿¡ ´ëÇÑ ÅëÇÕ ¼³¸í ½Ã½ºÅÛ |
¿µ¹®Á¦¸ñ(English Title) |
Integrated Explanation System for a Scalable Data based on SPARQL Results |
ÀúÀÚ(Author) |
Àü¸íÁß
¹ÚÇö±Ô
¹Ú¿µÅÃ
MyungJoong Jeon
HyunKyu Park
YoungTack Park
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 45 NO. 10 PP. 1004 ~ 1014 (2018. 10) |
Çѱ۳»¿ë (Korean Abstract) |
ÃÖ±Ù ´Ù¾çÇÑ QA½Ã½ºÅÛ ¹× Àü¹®°¡ ½Ã½ºÅÛ¿¡¼ ÁúÀÇ °á°ú¿¡ ´ëÇÑ ¼³¸íÀ» ÇÊ¿ä·Î ÇÏ´Â ¿ä±¸°¡ Áõ´ëµÇ°í ÀÖ´Ù. ÇÏÁö¸¸ ÇöÀç ¿¬±¸µÇ´Â ½Ã½ºÅÛ¿¡¼´Â ´ë¿ë·® ±â¹ÝÀÇ ÁúÀÇ Ã³¸®¿¡¸¸ ÁßÁ¡À» µÎ°í ÀÖ´Ù. µû¶ó¼ º» ³í¹®¿¡¼´Â ´ë¿ë·® µ¥ÀÌÅÍ ±â¹Ý ÁúÀÇ °á°ú¿¡ ´ëÇÑ Àΰú °ü°è¸¦ ¼³¸íÇÒ ¼ö ÀÖ´Â ÅëÇÕ ½Ã½ºÅÛÀ» Á¦¾ÈÇÑ´Ù. º» ½Ã½ºÅÛÀº ´ë¿ë·®ÀÇ Áö½Ä¿¡ ´ëÇÑ Ãß·ÐÀ» À§ÇØ ºÐ»ê ±ÔÄ¢ ±â¹Ý SWRL¿£ÁøÀ» »ç¿ëÇÑ´Ù. ±×¸®°í ±× Ãß·Ð ±Ù°Å¸¦ Àΰú °ü°è¿¡ ´ëÇÑ ±¸Á¶·Î Ç¥ÇöÇÒ ¼ö ÀÖ´Â ºÐ»ê °¡Á¤±â¹Ý Áø¸® °ü¸® ½Ã½ºÅÛ(Distributed ATMS)ÀÇ ÀÔ·ÂÀ¸·Î »ç¿ëÇÑ´Ù. ¸¶Áö¸·À¸·Î ´ë¿ë·® µ¥ÀÌÅÍ ±â¹ÝÀÇ SPARQL 󸮱âÀÎ SPARQLGX¸¦ »ç¿ëÇÏ¿© ÁúÀÇ ÇØ´äÀ» ±¸ÇÏ°í ±× ÇØ´ä¿¡ ´ëÇÑ Àΰú °ü°è¸¦ ¾Õ¼ ±¸ÃàµÈ ÀÇÁ¸ ±¸Á¶¸¦ ÂüÁ¶ÇÏ¿© ¼³¸íÇÑ´Ù. Á¦¾ÈÇÑ ÅëÇÕ ¼³¸í ½Ã½ºÅÛ¿¡ ´ëÇÑ Æò°¡´Â º¥Ä¡¸¶Å© µ¥ÀÌÅÍÀÎ LUBM(Lehigh University Benchmark)À» »ç¿ëÇßÀ¸¸ç, LUBM¿¡¼ Á¦°øÇÑ 14°³ÀÇ Å×½ºÆ® ÁúÀǹ®À» »ç¿ëÇÏ¿© ÁúÀÇ ÀÀ´ä ½Ã°£ ¹× ¼³¸í ½Ã°£¿¡ ´ëÇÑ ¼º´É Æò°¡¸¦ ÇÏ¿´´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
Recently, there has been an increasing demand for an explanation of query results in a variety of QA systems and expert systems. However, the systems being studied today only focus on the scalable query processing. Therefore, this paper proposes an integrated system that explains the causal relationship to the query results based on large volumes of retrievable data. The system uses a distributed rule-based SWRL engine for reasoning about large amounts of knowledge. And in this case uses evidence of reasoning as input for a distributed ATMS to express the structure of the causal relationship. Finally, after obtaining the answers using SPARQLGX, and a scalable SPARQL query processor, this system explains the evidence of answers using a reference to the previously established dependency structure. The evaluation of the proposed explanation system used the benchmark data(Lehigh University Benchmark) and used 14 test queries provided by the LUBM for evaluating the response time and explanation time in this case.
|
Å°¿öµå(Keyword) |
ÅëÇÕ ¼³¸í ½Ã½ºÅÛ
SPARQL
ºÐ»ê SWRL Ãß·Ð
ºÐ»ê °¡Á¤±â¹Ý Áø¸® °ü¸® ½Ã½ºÅÛ
integrated explanation system
SPARQL
distributed SWRL reasoner
distributed ATMS
SPARQ
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|