• 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 : 2 / 11 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) Æ®À­À» ÀÌ¿ëÇÑ ±×·¡ÇÁ¿Í ½Ã°è¿­ ±â¹Ý »çÀ̹ö °ø°Ý ¿¹Ãø
¿µ¹®Á¦¸ñ(English Title) Predicting of Cyber Attacks based on Graphs and Time-Series Using Tweets
ÀúÀÚ(Author) ÀÌÁØÇÏ   Junha Lee   ÇѺ¸¿µ   Boyoung Han   ±ÇÇõÀ±   Hyuk-Yoon Kwon  
¿ø¹®¼ö·Ïó(Citation) VOL 35 NO. 03 PP. 0003 ~ 0017 (2019. 12)
Çѱ۳»¿ë
(Korean Abstract)
º» ³í¹®¿¡¼­´Â Æ®À­À» ÀÌ¿ëÇÑ ±×·¡ÇÁ¿Í ½Ã°è¿­ ±â¹Ý »çÀ̹ö °ø°Ý ¿¹Ãø ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. º» ³í¹®Àº »çÀ̹ö °ø°Ý°ú ¿¬°üµÈ »ç¿ëÀÚµé °£ÀÇ °ü°è¸¦ Ç¥ÇöÇÑ ±×·¡ÇÁ µ¥ÀÌÅÍ¿Í ½Ã°£¿¡ µû¸¥ ±×µéÀÇ È°µ¿À» Ç¥ÇöÇÑ ½Ã°è¿­ µ¥ÀÌÅ͸¦ µ¿½Ã¿¡ È°¿ëÇÏ¿© »çÀ̹ö °ø°ÝÀ» ¿¹ÃøÇÑ ÃÖÃÊÀÇ ¿¬±¸ÀÌ´Ù. ¸ÕÀú, »çÀ̹ö °ø°Ý°ú ¿¬°üµÈ Å°¿öµå°¡ Æ÷ÇÔµÈ Æ®À­À» ÀÚÁÖ ÀÛ¼ºÇÏ´Â »ç¿ëÀÚµéÀ» ´ë»óÀ¸·Î »ç¿ëÀÚµé °£ÀÇ °ü°è¸¦ ±×·¡ÇÁ·Î ¸ðµ¨¸µÇÑ´Ù. À̸¦ ±â¹ÝÀ¸·Î »ç¿ëÀÚ °£ ±×·¡ÇÁ »óÀÇ ÀÎÁ¢¼ºÀ» °í·ÁÇÑ È¿°úÀûÀΠŬ·¯½ºÅ͸µ ±âÁØÀ» ¼±º°ÇÑ´Ù. ´ÙÀ½À¸·Î, »çÀ̹ö °ø°Ý¿¡ ´ëÇÑ ´º½º ±â»ç¸¦ ±â¹ÝÀ¸·Î Æ®À­ »ç¿ëÀÚµéÀÇ Æ®À­ ÀÛ¼º ÃßÀ̸¦ ½Ã°è¿­·Î ºÐ¼®ÇÑ´Ù. À̶§, »ç¿ëÀÚ ±×·ì °£ »çÀ̹ö °ø°Ý ÀüÈÄÀÇ Æ®À­ ÀÛ¼º ºóµµ¸¦ ºÐ¼®ÇÔÀ¸·Î½á, ±×·¡ÇÁ ºÐ¼®¿¡¼­ ¼±º°µÈ Ŭ·¯½ºÅ͸µ ±âÁØÀÌ »çÀ̹ö °ø°Ý ¿¹Ãø¿¡ È¿°úÀûÀÓÀ» º¸ÀδÙ. 2013-2018³â »çÀÌ¿¡ ¹ß»ýÇÑ ÃÑ 58°ÇÀÇ »çÀ̹ö °ø°Ý¿¡ ´ëÇÑ ºÐ¼®À» ¼öÇàÇÑ´Ù. ±¸Ã¼ÀûÀ¸·Î, »çÀ̹ö °ø°Ý°ú ¿¬°üµÈ Æ®À­À» ÀÚÁÖ ÀÛ¼ºÇÏ´Â 1,000¸íÀÇ »ç¿ëÀÚ Áß¿¡¼­ ÀÓÀÇ·Î ¼±Á¤µÈ 100¸íÀÇ »ç¿ëÀÚ ±×·ìÀÌ ÀÛ¼ºÇÑ Æ®À­¿¡ ºñÇÏ¿© ±×·¡ÇÁ »óÀÇ ÀÎÁ¢¼º¿¡ µû¸¥ Ŭ·¯½ºÅ͸µ ±âÁØ¿¡ µû¶ó ¼±º°µÈ 100¸íÀÇ »ç¿ëÀÚ ±×·ìÀÌ ÀÛ¼ºÇÑ Æ®À­ÀÌ »çÀ̹ö °ø°Ý ¿¹Ãø¿¡ ÃÖ´ë ¾à 18%¸¸Å­ ´õ È¿°úÀûÀÓÀ» º¸ÀδÙ.
¿µ¹®³»¿ë
(English Abstract)
In this paper, we propose a method for predicting of cyber attacks based on graphs and time series using tweets. This paper is the first research effort that utilizes both graph data representing the relationship between users who are related to cyber attacks and time-series data representing their activity according to the time. First, we perform a graph-based modeling for describing the relationship between users who frequently write tweets related to cyber attacks. Based on the graph, we identify the effective clustering criteria between users by taking into account their proximity on the graph. Next, we analyze the trend that the users write tweets in a time series based on news articles about cyber attacks. Here, by analyzing the frequency of tweets according to the groups of tweet users before and after cyber attacks, we show that the clustering criteria identified by the graph analysis is effective. We perform the analysis on a total of 58 cyber attacks between 2013 and 2018. Specifically, out of a total of 1,000 users who frequently write tweets related to cyber attacks, we show that tweets written by a group of 100 users, who are selected according to the proximity on the graph, are more effective for predicting cyber attacks by up to 18% compared to the tweets written by a group of 100 users, who are randomly selected.
Å°¿öµå(Keyword) predicting cyber attacks   »çÀ̹ö °ø°Ý ¿¹Ãø   ±×·¡ÇÁ ºÐ¼®   ½Ã°è¿­ ºÐ¼®   graph analysis   time-series analysis  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå