• 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) ÇÑ±Û Æ®À§ÅÍ»óÀÇ Äڷγª19 ¹é½Å ´ã·Ð¿¡ ´ëÇÑ 2´Ü°è µ¿Àû ÅäÇÈ ¸ðµ¨¸µ ºÐ¼®
¿µ¹®Á¦¸ñ(English Title) A Two-Phase Dynamic Topic Modeling Analysis of COVID-19 Vaccine Discourse in Korean Twitter
ÀúÀÚ(Author) ¼­Çϸ°   ¼­¿µ±Õ   Harin Seo   Young-Kyoon Suh  
¿ø¹®¼ö·Ïó(Citation) VOL 39 NO. 01 PP. 0017 ~ 0032 (2023. 04)
Çѱ۳»¿ë
(Korean Abstract)
º» ¿¬±¸´Â ¼Ò¼È ¹Ìµð¾î µ¥ÀÌÅÍ¿¡ ÀáÀçÇÏ´Â ±¹³» Äڷγª19 ¹é½Å ÀïÁ¡¿¡ ´ëÇÑ ½Ã°è¿­ ±â¹Ý ºÐ¼®À» À§ÇÑ ÃֽŠ2´Ü°è µ¿Àû ÅäÇÈ ¸ðµ¨¸µ(Dynamic Topic Modeling, DTM) ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. º» ¿¬±¸¸¦ À§ÇØ, ¿ì¸®´Â ±¹³» Äڷγª19 ¹é½Å Á¢Á¾ ½ÃÇà ¹× °èȹÀÌ º»°ÝÀûÀ¸·Î ½ÃÀÛµÈ 2021³â 2¿ùºÎÅÍ 2023³â 1¿ù±îÁö ÀÛ¼ºµÈ Æ®À§ÅÍ µ¥ÀÌÅ͸¦ ¼öÁýÇÏ¿© ºÐ¼®ÇÏ¿´´Ù. ±âÁ¸ DTM ±â¹ýÀº Ãß¼¼°¡ ±Þº¯ÇÏ´Â Äڷγª19 ¹é½Å Ư¼º»ó ÅäÇÈ ³» Å°¿öµåÀÇ ¿¬°ü¼ºÀ» ã±â ¾î·Á¿ö ÅäÇÈÀÇ ¼û°ÜÁø Àǹ̸¦ ºÐ¼®Çϱ⠾î·Æ´Ù´Â ¹®Á¦°¡ ÀÖ´Ù. ÀÌ·¯ÇÑ ºÐ¼®»óÀÇ ³­Á¦¸¦ ÇØ°áÇϱâ À§ÇØ, º» ¿¬±¸¿¡¼­´Â 1´Ü°è·Î Àüü Æ®À§ÅÍ µ¥ÀÌÅͼ¿¡ ´ëÇÑ °Å½ÃÀû DTMÀ» ¼öÇàÇÏ¿© ÅäÇÈ ³» Å°¿öµåµéÀÇ À¯»ç¼ºÀÌ º¯ÇÏ´Â ½ÃÁ¡À» ã°í, ÇØ´ç ½ÃÁ¡µéÀ» ±âÁØÀ¸·Î µ¥ÀÌÅͼÂÀ» À籸ÃàÇØ 2´Ü°è DTMÀ» ¼öÇàÇÑ´Ù. ±× °á°ú, ¿ì¸®´Â Á¦¾ÈÇÑ 2´Ü°è DTM¿¡¼­ ±âÁ¸ DTM´ëºñ ´õ ´Ù¾çÇÑ ÅäÇÈÀÌ È¿°úÀûÀ¸·Î ºÐ·ùµÇ´Â °ÍÀ» È®ÀÎÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
This study introduces a novel two-phase Dynamic Topic Modeling (DTM) technique for analyzing in time-series domestic COVID-19 vaccine-related issues latent in social media data. To this end, we collected and analyzed the Twitter data from February 2021, when starting in earnest COVID-19 vaccination implementation in Korea, to January 2023. The existing DTM technique revealed a problem in that it is difficult to analyze the hidden meaning of the topic because it is difficult to find the relevance of the keywords within the topic due to the rapidly changing trend of the COVID-19 vaccine. To resolve this challenge in analysis, in this study we conduct a macro DTM on the entire Twitter dataset to find the point at which the similarity of keywords within a topic changes, and proceed with the second phase DTM by reconstructing the dataset based on the point in time. As a result, we confirmed that the proposed two-phase DTM approach effectively identified a greater variety of topics than existing DTM.
Å°¿öµå(Keyword) Äڷγª19   ¼Ò¼È ¹Ìµð¾î   µ¿Àû ÅäÇÈ ¸ðµ¨¸µ   À̽´ ºÐ¼®   Äڷγª19 ¹é½Å   COVID-19   Social media   Dynamic topic modeling   Issue analysis   COVID-19 vaccine  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå