µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
µðÁöÅÐÇコÄÉ¾î ¼ºñ½º¿¡¼ ¹®ÀÚ¿ ¼öÁØÀÇ À¥ Æ®·¡ÇÈ °ø°ÝŽÁö¸¦ À§ÇÑ LSTM ¸ðµ¨ ¿¬±¸ |
¿µ¹®Á¦¸ñ(English Title) |
Character Level Web Traffic Attack Detection using LSTM Approaches in Digital Healthcare Service |
ÀúÀÚ(Author) |
±è¹®Çö
±è¿µÈ£
±è¿µ±¹
Mun-Hyeon Kim
Young-Ho Kim
Young-Kuk Kim
Á¤À¯Ã¤
Yuchae Jung
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 39 NO. 01 PP. 0096 ~ 0104 (2023. 04) |
Çѱ۳»¿ë (Korean Abstract) |
º´¿ø ³» Á¤º¸½Ã½ºÅÛÀº º´¿øÁ¤º¸½Ã½ºÅÛ, ó¹æÀü´Þ½Ã½ºÅÛ, ÀǷ῵»óÀúÀåÀü¼Û½Ã½ºÅÛ, ¿ø°ÝÁø·á ½Ã½ºÅÛ µîÀ¸·Î ±¸¼ºµÇ¾î ÀÖÀ¸¸ç °³ÀÎÀÇ ÀÇ·áÁ¤º¸ º¸È£¸¦ À§ÇØ ¿ÜºÎ¿Í ºÐ¸®µÈ ³»ºÎ¸ÁÀ¸·Î ¿¬°áµÇ¾î ÀÖ´Ù. ÃÖ±Ù¿¡´Â ´Ù¾çÇÑ µðÁöÅÐ ÇコÄÉ¾î ¼ºñ½º Á¦°ø ¹× Ŭ¶ó¿ìµå ±â¹Ý ÀÇ·áÁ¤º¸ ºÐ¼®À» À§ÇØ ¿ÜºÎ¸Á°ú Á¦ÇÑÀûÀ¸·Î ¿¬°áÀ» Çã¿ëÇÏ°í ÀÖ¾î º¸¾È Ãë¾àÁ¡ ´ëºñ¿Í °ü·ÃµÈ ¿ä±¸°¡ Áõ°¡ÇÏ°í ÀÖ´Ù. µðÁöÅÐÇコÄÉ¾î ¼ºñ½º¿¡ °ü·ÃµÈ º¸¾È Ãë¾àÁ¡ ¿¬±¸´Â Åë°èÀûÀÎ ¹æ¹ý¿¡ ±â¹ÝÇÑ ÀüÅëÀûÀÎ ¿¬±¸°¡ ´ëºÎºÐÀÌ¸ç µö·¯´× ±â¹ÝÀÇ ÃֽŠº¸¾ÈÀ§Çù ŽÁö¹æ¹ýÀº °ÅÀÇ Àû¿ëÇÏ°í ÀÖÁö ¾Ê´Ù. º» ¿¬±¸¿¡¼´Â µðÁöÅÐÇコÄÉ¾î ¼ºñ½º¿¡¼ ¹ß»ý °¡´ÉÇÑ ´Ù¾çÇÑ À¥°ø°Ý À¯ÇüÀ» Æ÷ÇÔÇÏ°í ÀÖ´Â 2Á¾ÀÇ µ¥ÀÌÅ͸¦ »ç¿ëÇØ ´Ù¾çÇÑ µö·¯´× ¸ðµ¨°ú ±â°èÇнÀ ¸ðµ¨À» ÈƷýÃÅ°°í º¸¾È°ø°Ý ŽÁö Á¤È®µµ¸¦ ºñ±³ ºÐ¼®ÇÏ¿´´Ù. µö·¯´× ¸ðµ¨ Áß LSTMÀÌ À¥ Æ®·¡ÇÈ¿¡¼ ¹ß»ý °¡´ÉÇÑ ´Ù¾çÇÑ º¸¾È °ø°ÝÀ» ¹®ÀÚ¿ ¼öÁØ¿¡¼ ³ôÀº Á¤È®µµ·Î ŽÁöÇÏ¿´´Ù. º» ¿¬±¸´Â À¥ Æ®·¡ÇÈ µ¥ÀÌÅ͸¦ ºñÁ¤»ó ŽÁöÇϴµ¥ ÃÖÀûÈ µÈ LSTM ¸ðµ¨À» Á¦¾ÈÇÔÀ¸·Î¼ ÇâÈÄ µðÁöÅÐ ÇコÄÉ¾î ¼ºñ½º¿¡¼ ¹ß»ý °¡´ÉÇÑ º¸¾ÈÀ§ÇùÀ» ³ôÀº Á¤È®µµ·Î ŽÁöÇÒ ¼ö ÀÖ´Â Â÷¼¼´ë Áö´ÉÇü º¸¾È°üÁ¦ ½Ã½ºÅÛ¿¡ È°¿ë °¡´ÉÇÔÀ» ½Ã»çÇÑ´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
In hospital, the information system consists of hospital information system (HIS), order communication system (OCS), and picture archiving and communication systems (PACS), telemedicine system and all of which are connected to an internal network for the protection of personal medical information. Recently, it allows limited connection with external networks to provide digital healthcare services and analyze cloud-based medical information, requiring preparation for security vulnerabilities. In this study, we preprocessed web traffic data at the character level, including normal and abnormal attack types, to detect various traffic attacks in digital healthcare services, and used them for training deep learning or machine learning models. Among the deep learning models, LSTM showed higher accuracy compared to CNN in detection of web traffic attack. This suggests that proposed LSTM model could be used for developing intelligent security monitoring system in the future, which could detect security threats in web services. |
Å°¿öµå(Keyword) |
¹®ÀÚ ÀνÄ
µö·¯´×
¾î¸°ÀÌ ¼Õ±Û¾¾
äÁ¡ ÀÚµ¿È
Character recognition
deep learning
scoring automation
ÁÖÁ¦¾î: LSTM
À¥Æ®·¡ÇÈ µ¥ÀÌÅÍ
º¸¾ÈÀ§Çù
µðÁöÅÐÇコÄÉ¾î ½Ã½ºÅÛ
LSTM
Web Traffic Data
Security Attack
Digital Healthcare System
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|