• 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¸Ã³¸®ÇÐȸ ³í¹®Áö > Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö ¼ÒÇÁÆ®¿þ¾î ¹× µ¥ÀÌÅÍ °øÇÐ

Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö ¼ÒÇÁÆ®¿þ¾î ¹× µ¥ÀÌÅÍ °øÇÐ

Current Result Document : 4 / 6 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) °Ë»ö ±â¹ÝÀÇ Áú¹®»ý¼º¿¡¼­ Áߺ¹ ¹æÁö¸¦ À§ÇÑ À¯»ç ÀÀ´ä °ËÃâ
¿µ¹®Á¦¸ñ(English Title) Detection of Similar Answers to Avoid Duplicate Question in Retrieval-based Automatic Question Generation
ÀúÀÚ(Author) ÃÖ¿ë¼®   ÀÌ°øÁÖ   Yong-Seok Choi   Kong Joo Lee  
¿ø¹®¼ö·Ïó(Citation) VOL 08 NO. 01 PP. 0027 ~ 0036 (2019. 01)
Çѱ۳»¿ë
(Korean Abstract)
º» ¿¬±¸´Â °Ë»ö ±â¹ÝÀÇ Áú¹® ÀÚµ¿ »ý¼º ½Ã½ºÅÛ¿¡¼­ »ç¿ëÀÚ°¡ ÀÌ¹Ì ´äº¯ÇÑ ³»¿ëÀ» ÀçÁú¹®ÇÏÁö ¾Êµµ·Ï »ç¿ëÀÚÀÇ ÀÀ´ä°ú À¯»çµµ°¡ ³ôÀº ÀÀ´äÀ» Áú¹®-µ¥ÀÌÅͺ£À̽º¿¡¼­ ã´Â ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. À¯»çµµ°¡ ³ô°Ô °ËÃâµÈ ÀÀ´äÀÇ Áú¹®Àº ÀÌ¹Ì »ç¿ëÀÚ°¡ ¾Æ´Â ³»¿ëÀÏ È®·üÀÌ ³ô±â ¶§¹®¿¡ Áú¹® È帱º¿¡¼­ Á¦°ÅÇÑ´Ù. À¯»ç ÀÀ´ä °ËÃâ¿¡´Â µÎ ÀÀ´ä°£ÀÇ µ¿ÀÏ ´Ü¾î, ¹Ù²ã¾²±â Ç¥Çö, ¹®Àå ³»¿ëÀ» ¸ðµÎ »ç¿ëÇÏ¿´´Ù. ¹Ù²ã¾²±â Ç¥ÇöÀº Åë°è±â¹ÝÀÇ ±â°è¹ø¿ª¿¡¼­ »ç¿ëÇÏ´Â ±¸Àý Å×À̺íÀ» »ç¿ëÇÏ¿© ±¸ÃàÇÏ¿´´Ù. ¹®Àå ³»¿ëÀº µÎ ¹®ÀåÀ» ÁÖÀÇ-±â¹Ý ÄÁº¼·ç¼Ç ½Å°æ¸ÁÀ¸·Î ¾ÐÃàÇÏ¿© À¯»çµµ¸¦ °è»êÇÏ¿´´Ù. Æò°¡¸¦ À§ÇØ ±¸ÃàÇÑ 100°³ÀÇ Æò°¡ ÀÀ´ä¿¡ Áú¹®-ÀÀ´ä µ¥ÀÌÅͺ£À̽º·ÎºÎÅÍ °¡Àå À¯»çÇÑ ÀÀ´äÀ» ÃßÃâÇؼ­ ¾òÀº °á°ú´Â MRR°ª 71%ÀÇ ¼º´ÉÀ» º¸¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
In this paper, we propose a method to find the most similar answer to the user¡®s response from the question-answer database in order to avoid generating a redundant question in retrieval-based automatic question generation system. As a question of the most similar answer to user¡¯s response may already be known to the user, the question should be removed from a set of question candidates. A similarity detector calculates a similarity between two answers by utilizing the same words, paraphrases, and sentential meanings. Paraphrases can be acquired by building a phrase table used in a statistical machine translation. A sentential meaning¡¯s similarity of two answers is calculated by an attention-based convolutional neural network. We evaluate the accuracy of the similarity detector on an evaluation set with 100 answers, and can get the 71% Mean Reciprocal Rank (MRR) score.
Å°¿öµå(Keyword) Áú¹®-ÀÀ´ä µ¥ÀÌÅͺ£À̽º   Áú¹® ÀÚµ¿ »ý¼º   À¯»ç ÀÀ´ä °ËÃâ   Question-Answer Database   Automatic Question Generation   Similar Answer Detection  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå