• 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) ÆòÁ¡ Á¤±ÔÈ­¸¦ ÀÌ¿ëÇÏ¿© »ç¿ëÀÚ Æò°¡ ¼ºÇâÀ» ¹Ý¿µÇÑ ¿µÈ­ Ãßõ ¹æ¹ý
¿µ¹®Á¦¸ñ(English Title) Movie Recommendation Method Using Score Normalization Based on User Rating Tendency
ÀúÀÚ(Author) ±èÇö°æ   ±èÇöÁø   ¹Ú»óÇö   Hyunkyung Kim   Hyunjin Kim   Sanghyun Park  
¿ø¹®¼ö·Ïó(Citation) VOL 32 NO. 02 PP. 0020 ~ 0031 (2016. 08)
Çѱ۳»¿ë
(Korean Abstract)
±âÁ¸ÀÇ »ç¿ëÀÚ ±â¹Ý Ãßõ ¹æ¹ýÀ» ÀÌ¿ëÇÑ ¿µÈ­ Ãßõ ½Ã½ºÅÛ¿¡¼­´Â ´ë°³ ´Ù¸¥ »ç¿ëÀÚµéÀÇ ÆòÁ¡À» ±â¹ÝÀ¸·Î ¸ñÇ¥ »ç¿ëÀÚÀÇ ÆòÁ¡À» ¿¹ÃøÇÏ´Â µ¥¿¡ ÀÌ¿ëÇÏ¿´Áö¸¸ »ç¿ëÀÚ °³°³ÀÎÀÇ Æò°¡ ¼ºÇâÀº ¹Ý¿µÇÏÁö ¾Ê¾Æ ÆòÁ¡ µ¥ÀÌÅÍÀÇ °´°ü¼ºÀ» È®º¸Çϱ⿡´Â ¾î·Á¿î Á¡ÀÌ ÀÖ¾ú´Ù. º» ³í¹®¿¡¼­´Â ±âÁ¸ÀÇ »ç¿ëÀÚ ±â¹Ý Ãßõ ¹æ¹ý°ú Ç׸ñ ±â¹Ý Ãßõ ¹æ¹ýÀ» ¹ÙÅÁÀ¸·Î ÇÑ Ç׸ñ °£ ¼±È£µµ Â÷À̸¦ ÀÌ¿ëÇÑ Ãßõ ¹æ¹ýÀ» Åä´ë·Î »ç¿ëÀÚ °³ÀÎÀÇ Æò°¡ ¼ºÇâÀ» ¹Ý¿µÇÑ »õ·Î¿î ¿µÈ­ Ãßõ ½Ã½ºÅÛÀ» Á¦¾ÈÇÑ´Ù. ¸¹Àº »ç¿ëÀÚµéÀÇ ÆòÁ¡ µ¥ÀÌÅÍ°¡ »ç¿ëÀÚÀÇ ¼ºÇâ¿¡ µû¶ó Ä¡¿ìÃÄ ÀÖ¾î ´Ù¸¥ »ç¿ëÀÚÀÇ ÆòÁ¡ ¿¹Ãø¿¡ ÀÌ¿ëµÇ±â¿¡´Â ´Ù¼Ò ¾î·Á¿î Á¡ÀÌ ÀÖ¾ú´Ù. µû¶ó¼­ »ç¿ëÀÚµéÀÇ Æò°¡ ¼ºÇâÀ» ¹ÙÅÁÀ¸·Î µ¥ÀÌÅ͸¦ Á¤±ÔÈ­ÇÏ¿´°í Ç׸ñ °£ ¼±È£µµ Â÷À̸¦ ÀÌ¿ëÇÏ¿© ÆòÁ¡À» ¿¹ÃøÇÏ´Â ½Ã½ºÅÛÀ» ±¸ÇöÇÏ¿´´Ù. ½ÇÇè °á°ú Á¦¾ÈÇÑ ½Ã½ºÅÛÀº ±âÁ¸ÀÇ ½Ã½ºÅÛ¿¡ ºñÇØ ÃßõÀÇ Á¤È®µµ°¡ Çâ»óµÇ¾ú´Ù. µû¶ó¼­ º» ¿¬±¸ÀÇ Á¦¾È ¹æ¹ýÀº »ç¿ëÀÚÀÇ ÆòÁ¡ °áÁ¤ ¼ºÇâÀ» ¹Ý¿µÇÔÀ¸·Î½á ´Ù¾çÇÑ ÄÜÅÙÃ÷¿¡ ´ëÇÑ »ç¿ëÀÚÀÇ Æò°¡¸¦ º¸´Ù Á¤È®ÇÏ°Ô ¿¹ÃøÇÏ¿© »ç¿ëÀÚ °³Àο¡ ¸Â´Â ¿µÈ­ ÃßõÀ» °¡´ÉÇÏ°Ô ÇÒ °ÍÀ¸·Î ±â´ëµÈ´Ù.
¿µ¹®³»¿ë
(English Abstract)
Existing movie recommender systems generally use rating data of other users to predict the rating of target user. However, it is hardly possible to guarantee the objectivity of rating data since the rating tendency of individual user is not considered. In this paper we propose a new recommendation method which took into account rating tendency of each individual user using score normalization based on rating difference between items. We have found many users have biased rating tendency and their rating data was affected. So we have normalized those rating data to get better prediction results. The results of experiments indicate that the proposed system has relatively improved performance compared to the previous recommender system in terms of prediction accuracy. Consequently, the proposed system is expected to enable improved movie recommendation for each individual by weighing rating tendency using score normalization.
Å°¿öµå(Keyword) Ãßõ ½Ã½ºÅÛ   µ¥ÀÌÅÍ ¸¶ÀÌ´×   Çù¾÷ ÇÊÅ͸µ   »ç¿ëÀÚ ¼ºÇâ   ÆòÁ¡ Á¤±ÔÈ­   Recommender System   Data Mining   Collaborative Filtering   Rating Tendency   Score Normalization  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå