• 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 :

ÇѱÛÁ¦¸ñ(Korean Title) ¿ÀÅä ÀÎÄÚ´õ ±â¹Ý Ãßõ ½Ã½ºÅÛÀ» À§ÇÑ ÀáÀç Ç¥Çö ÇнÀ ¹æ¹ý
¿µ¹®Á¦¸ñ(English Title) Latent Representation Learning for Autoencoder-based Top-K Recommender System
ÀúÀÚ(Author) ¹Úµ¿¹Î   °­ÁØÇõ   ÀÌÀç±æ   Dongmin Park   Junhyeok Kang   Jae-Gil Lee  
¿ø¹®¼ö·Ïó(Citation) VOL 47 NO. 02 PP. 0207 ~ 0215 (2020. 02)
Çѱ۳»¿ë
(Korean Abstract)
¿Â¶óÀÎ »óÀÇ »óÇ°ÀÇ ¼ö°¡ ±âÇÏ ±Þ¼öÀûÀ¸·Î Áõ°¡ÇÔ¿¡ µû¶ó °í°´ÀÌ ½º½º·Î ¿øÇÏ´Â »óÇ°À» ã´Â °ÍÀÌ ¾î·Á¿öÁ³´Ù. ÀûÀýÇÑ »óÇ°ÀÇ ÃßõÀº °í°´ÀÇ ÀáÀçÀû ¼ö¿ä¸¦ ¸¸Á·½ÃÅ°°í ÆǸÅÀÚÀÇ ÀÌÀ±À» Áõ´ë½ÃÅ°±â¿¡ ±× Á߿伺ÀÌ »ó´çÈ÷ Å©´Ù. ÃÖ±Ù¿¡´Â Àΰø½Å°æ¸ÁÀ» È°¿ëÇÑ Â÷¿ø Ãà¼Ò ±â¹ýÀÎ ¿ÀÅä ÀÎÄÚ´õ ±â¹ÝÀÇ Çù¾÷ ÇÊÅ͸µ ¹æ¹ýÀÌ ¼º´É ¸é¿¡¼­ µÎ°¢À» ³ªÅ¸³»¾ú´Ù. ÇÏÁö¸¸, ¿ÀÅä ÀÎÄÚ´õÀÇ ÀáÀç Ç¥Çö ºÐÆ÷ Á¶Á¤À» ÅëÇØ Ãßõ ¼º´ÉÀ» Çâ»ó½ÃÅ°´Â ¹æ¹ýÀº ¾ÆÁ÷ ¸¹ÀÌ ¿¬±¸µÇÁö ¾Ê¾Ò´Ù. º» ¿¬±¸¿¡¼­´Â ¿ÀÅä ÀÎÄÚ´õ ±â¹Ý Çù¾÷ ÇÊÅ͸µ ¹æ¹ý¿¡ °áÇÕµÇ¾î »óÇ° Ãßõ ¼º´ÉÀ» ´õ¿í Çâ»ó½ÃÅ°´Â ¹ÐÁý ÀáÀç Ç¥Çö ÇнÀ¹æ¹ý (DenseLR)À» Á¦¾ÈÇÑ´Ù. º» ¿¬±¸ÀÇ ÇÙ½É ¾ÆÀ̵ð¾î´Â À¯Àú ±¸¸Å Á¤º¸ º¤Å͵éÀÇ ÀáÀç Ç¥ÇöÀ» È¿°úÀûÀ¸·Î ¹ÐÁý ½ÃÅ´¿¡ µû¶ó Ãà¼Ò Â÷¿ø¿¡¼­ÀÇ Çù¾÷ ÇÊÅ͸µ È¿°ú¸¦ °­È­ÇÏ´Â °ÍÀÌ´Ù. 3°¡Áö ½ÇÁ¦ ±¸¸Å µ¥ÀÌÅÍ ¼Â¿¡ ´ëÇØ ±âÁ¸ ÃÖ÷´Ü ¿¬±¸µé°ú ¼º´É ºñ±³½ÇÇèÀ» ÁøÇàÇÑ °á°ú Á¦¾È ¹æ¹ýÀÌ ¸ðµç µ¥ÀÌÅÍ ¼Â¿¡ ´ëÇØ °¡Àå ³ôÀº ¼º´ÉÀ» º¸¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
As the number of products on the Internet is growing exponentially, it becomes more difficult for customers to choose the product they want. Many researchers have been actively making efforts to develop appropriate recommender systems that satisfy the potential demand of the customer and increase the profit of the seller. Recently, collaborative filtering methods based on an autoencoder have shown high performance. However, little attention has been paid for improving the recommendation performance by changing the distribution of latent representation. In this paper, we propose the Dense Latent Representation learning method (DenseLR) which is combined with the autoencoderbased collaborative filtering method to further improve product recommendation performance. The key idea of the DenseLR is to tighten collaborative filtering effects on the latent space by effectively densifying the latent representations of user (or item) rating vectors. In performance compariso experiments on three real-world datasets, DenseLR showed the highest recommendation performance for all datasets. Furthermore, DenseLR can be flexibly combined with a wide range of autoencoderbased CF models, and we empirically validated the improvement of the f1@k score ranging from 4.6% to 23.7%.
Å°¿öµå(Keyword) Ç¥Çö ÇнÀ   recommender system   collaborative filtering   autoencoder   representation learning  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå