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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)

Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)

Current Result Document : 7 / 8

ÇѱÛÁ¦¸ñ(Korean Title) Ãʱ⠼ҷ® µ¥ÀÌÅÍ¿Í RNNÀ» È°¿ëÇÑ ·ç¸Ó ÀüÆÄ ÃßÀû ±â¹ý
¿µ¹®Á¦¸ñ(English Title) Initial Small Data Reveal Rumor Traits via Recurrent Neural Networks
ÀúÀÚ(Author) ±Ç¼¼Á¤   Â÷¹Ì¿µ   Sejeong Kwon   Meeyoung Cha  
¿ø¹®¼ö·Ïó(Citation) VOL 44 NO. 07 PP. 0680 ~ 0685 (2017. 07)
Çѱ۳»¿ë
(Korean Abstract)
¿Â¶óÀÎ ¼Ò¼È¹Ìµð¾îÀÇ µîÀåÀ¸·Î ¹æ´ëÇÑ »ç¿ëÀÚ µ¥ÀÌÅÍ°¡ ¼öÁýµÇ°í ÀÌ´Â ·ç¸ÓÀÇ Å½Áö¿Í °°Àº º¹ÀâÇÏ°í µµÀüÀûÀÎ »çȸ ¹®Á¦¸¦ ÀÚ·á ±â¹Ý ±â¹ýÀ¸·Î ÇØ°áÇÒ ¼ö ÀÖ°Ô²û ÇÑ´Ù. ÃÖ±Ù µö·¯´× ±â¹Ý ¸ðµ¨µéÀÌ ÀÌ·¯ÇÑ ¹®Á¦¸¦ ÇØ°áÇϱâ À§ÇÑ ºü¸£°í Á¤È®ÇÑ ±â¹ý ÁßÀÇ Çϳª·Î¼­ ¼Ò°³µÇ¾ú´Ù. ÇÏÁö¸¸ ±âÁ¸¿¡ Á¦½ÃµÈ ¸ðµ¨µéÀº ÀüÆÄ Á¾·á ÈÄ ÀÛµ¿Çϰųª ¿À·£ °üÂû±â°£À» ÇÊ¿ä·Î ÇÏ¿© È°¿ë¼ºÀÌ Á¦ÇѵȴÙ. ÀÌ ¿¬±¸¿¡¼­´Â Ãʱ⠼ҷ®µ¥ÀÌÅ͸¸À» È°¿ëÇÏ´Â recurrent neural networks (RNNs) ±â¹ÝÀÇ ºü¸¥ ·ç¸Ó ºÐ·ù ¾Ë°í¸®ÁòÀ» Á¦¾ÈÇÑ´Ù. Á¦½ÃµÈ ¸ðµ¨Àº ¼Ò¼È¹Ìµð¾î ½ºÆ®¸²À» ½Ã°è¿­ ÀÚ·á·Î º¯È¯ÇÏ¿© »ç¿ëÇϸç, ÀÌ ¶§ ½Ã°è¿­ µ¥ÀÌÅÍ´Â Æȷοö ¼ö¿Í °°ÀÌ Á¤º¸ ÀüÆÄÀÚ °ü·Ã Á¤º¸´Â ¹°·Ð ÁÖ¾îÁø ÄÁÅÙÃ÷¿¡¼­ Ãß·ÐÇÑ ¾ð¾î½É¸®ÇÐÀû °¨¼ºÀÇ Á¡¼ö·Î ±¸¼ºµÈ´Ù. ¼ö¹é¸¸ÀÇ Æ®À­À» Æ÷ÇÔÇÏ´Â 498°³ÀÇ ½ÇÁ¦ ·ç¸Ó ¹× 494°³ÀÇ ºñ·ç¸Ó »ç·Ê ºÐ¼®À» ÅëÇØ ÀÌ ¿¬±¸´Â Á¦¾ÈÇÏ´Â RNN ±â¹Ý ¸ðµ¨ÀÌ Ãʱâ 30°³ÀÇ Æ®À­ ¸¸À¸·Îµµ (Ãʱ⠼ö½Ã°£) 0.74 F1ÀÇ ³ôÀº ¼º´ÉÀ» º¸ÀÓÀ» È®ÀÎÇÑ´Ù. ÀÌ·¯ÇÑ °á°ú´Â ½ÇÁ¦ ÀÀ¿ë°¡´ÉÇÑ ¼öÁØÀÇ ºü¸£°í È¿À²ÀûÀÎ ·ç¸Ó ºÐ·ù ¾Ë°í¸®Áò °³¹ßÀÇ Ãʼ®ÀÌ µÈ´Ù.
¿µ¹®³»¿ë
(English Abstract)
The emergence of online media and their data has enabled data-driven methods to solve challenging and complex tasks such as rumor classification problems. Recently, deep learning based models have been shown as one of the fastest and the most accurate algorithms to solve such problems. These new models, however, either rely on complete data or several days-worth of data, limiting their applicability in real time. In this study, we go beyond this limit and test the possibility of super early rumor detection via recurrent neural networks (RNNs). Our model takes in social media streams as time series input, along with basic meta-information about the rumongers including the follower count and the psycholinguistic traits of rumor content itself. Based on analyzing millions of social media posts on 498 real rumors and 494 non-rumor events, our RNN-based model detected rumors with only 30 initial posts (i.e., within a few hours of rumor circulation) with remarkable F1 score of 0.74. This finding widens the scope of new possibilities for building a fast and efficient rumor detection system.
Å°¿öµå(Keyword) ·ç¸Ó   µö·¯´×   ½Ã°è¿­   ¾ð¾îÀû Ư¡   »ç¿ëÀÚ Æ¯Â¡   ºÐ·ù±â   Rumor   deep learning   time series   user traits   linguistic traits   classification  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå