µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)
Current Result Document : 1 / 1
ÇѱÛÁ¦¸ñ(Korean Title) |
±â¹ýÀ» ÀÌ¿ëÇÑ Àü¹®°¡ Ãßõ ±â¹ý |
¿µ¹®Á¦¸ñ(English Title) |
An Expert Recommendation Technique using Hybrid Collaborative Filtering in SNS |
ÀúÀÚ(Author) |
½Å¿µ¼º
¿À¿µ¸¸
¿Àº´¼®
±èÇüÀÏ ÀåÀç¿ì
Young-Sung Shin
Young-Man Oh
Byeong-Seok Oh
Hyeong-il Kim
Jae-woo Chang
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 28 NO. 02 PP. 0003 ~ 0017 (2012. 08) |
Çѱ۳»¿ë (Korean Abstract) |
ÃÖ±Ù ´Ù¾çÇÑ Á÷¾÷À» °¡Áø SNS »ç¿ëÀÚ°¡ Áõ°¡ÇÔ¿¡ µû¶ó, SNS »ç¿ëÀÚµéÀº Àü¹®°¡ °£ Çù¾÷ ¹× ±â¼úÀû ÀÇ»ç¼ÒÅëÀ» À§ÇÑ Àü¹®°¡ Ãßõ ±â´ÉÀ» ¿ä±¸ÇÏ°í ÀÖ´Ù. ÇÏÁö¸¸ ±âÁ¸ SNS¿¡¼ Á¦°øÇÏ°í ÀÖ´Â Çù¾÷Àû ¿©°ú ±â¹ýÀº Àü¹®°¡ Ãßõ ¼ºñ½º¸¦ È¿À²ÀûÀ¸·Î Á¦°øÇÏÁö ¸øÇÑ´Ù. µû¶ó¼ ±âÁ¸ÀÇ Çù¾÷Àû ¿©°ú ±â¹ýÀÇ ´ÜÁ¡À» ±Øº¹ÇÏ°í ´Ù¾çÇÑ ÃßõÀ» ¼öÇàÇÒ ¼ö ÀÖ´Â ÇÏÀ̺긮µå Çù¾÷Àû ¿©°ú ±â¹ýÀÌ °³¹ßµÇ¾ú´Ù. ±× °¡¿îµ¥ Content-boosted Çù¾÷Àû ¿©°ú ±â¹ýÀº ´Ù¾çÇÑ ¿¹Ãø ¾Ë°í¸®ÁòÀ» Á¦½ÃÇÏ¿©, È¿°úÀûÀÎ ÃßõÀ» ¼öÇàÇÒ ¼ö ÀÖµµ·Ï Áö¿øÇÑ´Ù. ±×·¯³ª ¸íÈ®ÇÑ °è»êÁ¶°ÇÀÌ Á¦½ÃµÇÁö ¸øÇÏ´Â °æ¿ì ¾ÆÀÌÅÛ ¹× »ç¿ëÀÚ À¯»çµµ °è»êÀ» ¼öÇàÇÒ ¼ö ¾ø´Â ´ÜÁ¡ÀÌ Á¸ÀçÇÑ´Ù. µû¶ó¼ º» ³í¹®¿¡¼´Â Content-boosted Çù¾÷Àû ¿©°ú ±â¹ýÀÇ ´ÜÁ¡À» ÇØ°áÇÏ´Â ÇÏÀ̺긮µå Çù¾÷Àû ¿©°ú±â¹ýÀ» ÀÌ¿ëÇÑ »õ·Î¿î Àü¹®°¡ Ãßõ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. ¶ÇÇÑ, À̸¦ ÀÌ¿ëÇÏ¿© SNS¿¡¼ÀÇ Àü¹®°¡ Ãßõ ½Ã½ºÅÛÀ» ¼³°èÇÑ´Ù. ¸¶Áö¸·À¸·Î ¼º´ÉÆò°¡¸¦ ÅëÇØ Á¦¾ÈÇÏ´Â ¾Ë°í¸®ÁòÀÌ ¼ºñ½º Á¤È®µµ Ãø¸é¿¡¼ ±âÁ¸ ±â¹ýº¸´Ù ¿ì¼öÇÔÀ» º¸ÀδÙ.
|
¿µ¹®³»¿ë (English Abstract) |
Recently, because wide range of professionals utilize Social Network Service(SNS), the SNS users need an expert recommendation service to enable their cooperation or technical communication. But, the existing collaborative filtering technique cannot support an expert recommendation service efficiently. Therefore, hybrid collaborative filtering techniques have been developed not only to improve the existing collaborative filtering method but also to perform various recommendations. Among them, content?boosted collaborative filtering provides various prediction algorithms which enable to support effective recommendations. However, the content?boosted collaborative filtering method cannot calculates the similarity of items or users when the definition of calculation condition is not clearly provided. Therefore, to solve the problem of existing content?boosted collaborative filtering method, we propose a hybrid collaborative filtering based expert recommendation scheme. Also, we design the expert recommendation system by using proposed method in SNS. Finally, from a performance analysis we show that our proposed algorithm outperforms the existing method in terms of service accuracy.
|
Å°¿öµå(Keyword) |
Çù¾÷Àû ¿©°ú
Àü¹®°¡ Ãßõ
¼Ò¼È ³×Æ®¿öÅ© ¼ºñ½º
Collaborative Filtering
Expert Recommendation
Social Network Service
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|