Á¤º¸°úÇÐȸ ³í¹®Áö A : ½Ã½ºÅÛ ¹× ÀÌ·Ð
Current Result Document : 2 / 2
ÇѱÛÁ¦¸ñ(Korean Title) |
ºÐ»ê ¸Þ¸ð¸® ´ÙÁß ÇÁ·Î¼¼¼ ȯ°æ¿¡¼ º´·Ä Àǹ̸Á ¸ðµ¨ÀÇ ±¸Çö |
¿µ¹®Á¦¸ñ(English Title) |
Implementation of Parallel Semantic Network Model on Distributed Memory Multiprocessor Environment |
ÀúÀÚ(Author) |
Á¤»óÈ
¹ÚÁøÈ£
Sanghwa Chung
Jinho Park
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 24 NO. 07 PP. 0675 ~ 0685 (1997. 07) |
Çѱ۳»¿ë (Korean Abstract) |
ÀΰøÁö´É ±â¼úÀÌ ±Ô¸ð°¡ Å« ½ÇÁ¦ÀûÀÎ ¹®Á¦¿¡ Àû¿ëµÊ¿¡ µû¶ó °úµµÇÑ Å½»ö°ú °è»ê·®À» ÁÙÀ̱â À§ÇÑ ¹æ¹ýÀÌ ÀΰøÁö´É ¿¬±¸ÀÇ Áß¿äÇÑ ¹®Á¦°¡ µÇ¾ú´Ù. ÀÌ·¯ÇÑ ¹®Á¦¿¡ ´ëÇÑ ±Ùº»ÀûÀÎ ÇØ°áÃ¥Àº ÀΰøÁö´É ¹®Á¦¿¡ º´·Äó¸® ±â¹ýÀ» Àû¿ëÇÏ´Â °ÍÀÌ´Ù. º» ³í¹®¿¡¼´Â semantic network»ó¿¡¼ È¿°úÀûÀÎ Ãß·ÐÀ» ÇÒ ¼ö ÀÖ´Â medium-grain ºÐ»ê ¸Þ¸ð¸® ´ÙÁß Ã³¸® ¸ðµ¨¿¡ °üÇÑ ¿¬±¸¸¦ ÇÏ¿´´Ù. ¸ðµ¨¿¡¼ °¡Àå Áß¿äÇÑ ºÎºÐÀ» ÇÁ·Î¼¼¼ÀÇ ±¸Á¶¿Í ¸ðµ¨ÀÇ ¸í·ÉÁýÇÕÀÌ´Ù. ¸í·ÉÁýÇÕÀº 20¿©°³ÀÇ °í¼öÁØ ¸í·ÉÀ¸·Î Á¤ÀÇÇÏ¿´°í ÀÌ¿¡ ÀûÇÕÇÑ ÇÁ·Î¼¼¼ ±¸Á¶¸¦ Á¦½ÃÇÏ¿´´Ù. ¶ÇÇÑ Á¦½ÃµÈ ¸ðµ¨¿¡ ´ëÇÏ¿© ½Ã¹Ä·¹À̼ÇÀ» ÅëÇÏ¿© ¼º´ÉÀ» ÃøÁ¤, ºÐ¼®ÇÏ¿© medium-grain ºÐ»ê ¸Þ¸ð¸® ´ÙÁß Ã³¸® ¸ðµ¨ÀÇ À¯È¿¼ºÀ» ÀÔÁõÇÏ¿´´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
As the artificial intelligence(AI) technologies have been applied to practical large scale applications, it has become an serious problem of AI research to reduce excessive search and computation. A fundamental solution to this problem would be applying parallel processing technology to the AI problems. In this paper we propose a medium-grain distributed memory multiprocessor system model which can do inference efficiently on semantic networks. In the model design, the most important parts are the processor architecture and the instruction set of the model. A discrete-event simulator is developed for the model, and several experiments are done on various inference mechanism to test and analyze the performance of the model. The result shows that the proposed model is efficient and effective with proper semantic node allocation schemes.
|
Å°¿öµå(Keyword) |
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|