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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ÇÐȸÁö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ÇÐȸÁö > µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) ÀüÀå »óȲÀ» ¹¦»çÇÏ´Â °¡¼³µéÀ» À§ÇÑ ÁØÁöµµ ±ºÁýÈ­ ¾Ë°í¸®Áò
¿µ¹®Á¦¸ñ(English Title) A Semi-supervised Clustering Algorithm for Hypotheses that Describe Battlefield Situations
ÀúÀÚ(Author) Á¶Çö¼ö   ÁÖÇöÁø   ¾ÈÁ¾Ã¶   Áø¼Ò¿¬   ½ÅÀ¯°æ   ½Å±âÁ¤   Hyeonsoo Jo   Hyunjin Choo   Jong-Chul Ahn   Soyeon Jin   Yukyung Shin. Kijung Shin  
¿ø¹®¼ö·Ïó(Citation) VOL 39 NO. 01 PP. 0072 ~ 0084 (2023. 04)
Çѱ۳»¿ë
(Korean Abstract)
ÃÖ±Ù ¿¬±¸¿¡¼­´Â ÀüÀå »óȲ ºÐ¼®¿ë ÀΰøÁö´É °³¹ßÀ» À§ÇÑ µ¥ÀÌÅͼ »ý¼º ¹æ¹ýÀÌ Á¦¾ÈµÇ¾úÀ¸¸ç, ÀÌ µ¥ÀÌÅÍ ¼ÂµéÀº ÀΰøÁö´ÉÀ» È°¿ëÇÑ ±º Âü¸ð ±â¼úÀÇ °³¹ßÀ» °¡¼ÓÈ­ÇÏ°í ÀÖ´Ù. ÁöÈÖ°üÀÇ ÀÇ»ç °áÁ¤À» È¿°úÀûÀ¸·Î Áö¿øÇÏ ±â À§Çؼ­´Â ÀüÀå »óȲÀ» Ç¥ÇöÇÏ´Â °¡¼³ ³»ÀÇ ´Ù¾çÇÑ ÀüÀå Áö½Ä ¿ä¼ÒµéÀ» ºÐ¼®ÇÏ°í, ¿¬°ü¼ºÀÌ ³ôÀº °¡¼³µéÀ» ±ºÁýÈ­ÇÒÇÊ¿ä°¡ÀÖ´Ù. ÇÏÁö¸¸°¡¼³³»¿¡´Â¼­·Î´Ù¸¥Æ¯¼ºÀ»°®´Â´Ù¾çÇÑÀüÀåÁö½Ä¿ä¼ÒµéÀ̺¹ÀâÇÏ°Ô¾ôÇô ÀÖ¾î, ÀÌ¿¡ ´ëÇÑ ÃæºÐÇÑ °í·Á ¾øÀÌ ´Ü¼øÈ÷ ±âÁ¸ÀÇ ±ºÁýÈ­ ¹æ¹ýÀ» Àû¿ëÇÏ´Â °Í¸¸À¸·Î´Â ¼º´ÉÀÇ ÇÑ°è°¡ ÀÖ´Ù. º» ³í¹®¿¡¼­´Â °¡¼³ ³»ÀÇ ´Ù¾çÇÑ ÀüÀå Áö½Ä ¿ä¼Òµé »çÀÌÀÇ °ü°è¸¦ È¿°úÀûÀ¸·Î ¹Ý¿µÇÏ¿© °¡¼³ °£ÀÇ À¯»çµµ¸¦ ÃßÁ¤ÇÏ°í, À̸¦ ±â¹ÝÀ¸·Î À¯»çÇÑ °¡¼³À» ±ºÁýÈ­ÇÏ´Â ÁØÁöµµ ¾Ë°í¸®ÁòÀ» Á¦¾ÈÇÏ¿´´Ù. ¶ÇÇÑ, 9°³ ÁÖÁ¦¿¡ ´ëÇØ »ý¼ºµÈ ÀüÀå »óȲ °¡¼³ µ¥ÀÌÅͼÂÀ» È°¿ëÇÑ ½ÇÇèÀ» ÅëÇØ, Á¦¾ÈÇÑ ¾Ë°í¸®ÁòÀÌ ±âÁ¸ÀÇ ¹æ¹ýµéº¸´Ù ÀÏ°üµÇ°Ô ¿ì¼öÇÑ ¼º´ÉÀ» º¸ÀÌ´Â °ÍÀ» È®ÀÎÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
In a recent work, a hypothesis-generation method for AI-based battlefield analyses was proposed, and it has accelerated the development of AI-based technologies for supporting military staff. In order to effectively support commanders¡¯ decision-making, it is necessary to analyze various battlefield knowledge elements in hypotheses describing the battlefield situations and cluster highly related hypotheses. However, in each battlefield situation hypothesis, various battlefield knowledge elements with different characteristics are intricately intertwined, making it sub-optimal to cluster hypotheses with existing clustering methods. In this paper, we propose a semi-supervised algorithm that accurately estimates the similarity between hypotheses by effectively utilizing various battlefield knowledge elements in the hypothesis and clusters similar hypotheses. Furthermore, using datasets generated for nine topics, we experimentally demonstrate that the proposed algorithm consistently and significantly outperformed existing clustering methods.
Å°¿öµå(Keyword) ÀüÀå»óȲ ºÐ¼®   ÀüÀå»óȲ ºÐ·ù   ±ºÁýÈ­ ¾Ë°í¸®Áò   ÀΰøÁö´É   Battlefield Analysis   Battlefield Classification   Clustering Algorithm   Artificial Intelligence  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå