Á¤º¸°úÇÐȸ³í¹®Áö (Journal of KIISE)
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
°íÀ¯¸í»ç ±âȣȸ¦ ÅëÇÑ ½Å°æ¸Á±â¹Ý ÇÑ¿µ ¹ø¿ª |
¿µ¹®Á¦¸ñ(English Title) |
Kor-Eng NMT using Symbolization of Proper Nouns |
ÀúÀÚ(Author) |
±è¸íÁø
³²ÁØ¿µ
Á¤Èñ¼®
ÃÖÈñ¿
Myungjin Kim
Junyeong Nam
Heeseok Jung
Heeyoul Choi
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 48 NO. 10 PP. 1084 ~ 1089 (2021. 10) |
Çѱ۳»¿ë (Korean Abstract) |
½Å°æ ±â°è ¹ø¿ª ºÐ¾ß´Â µö·¯´×ÀÇ ¹ßÀü°ú ÇÔ²² ¼º´ÉÀÌ ¹ßÀüÇÏ°í ÀÖÁö¸¸, À̸§, ½ÅÁ¶¾î, ƯÁ¤ ±×·ì ³»¿¡¼¸¸ Åë¿ëµÇ´Â ´Ü¾î µî°ú °°ÀÌ °íÀ¯¸í»çµéÀÌ µé¾î°£ ¹®ÀåÀÇ ¹ø¿ªÀÌ Á¤È®ÇÏÁö ¾ÊÀº °æ¿ìµéÀÌ ÀÖ´Ù. º» ³í¹®Àº °íÀ¯¸í»ç°¡ µé¾î°£ ¹®ÀåÀÇ ¹ø¿ª ¼º´É °³¼±À» À§ÇØ ÃÖ±Ù Á¦¾ÈµÈ ¹ø¿ª ¸ðµ¨ÀÎ Transformer Model¿¡ Ãß°¡ÀûÀ¸·Î ÇÑ¿µ °íÀ¯¸í»ç »çÀü°ú °íÀ¯¸í»ç ±âÈ£È ¹æ½ÄÀ» »ç¿ëÇÑ´Ù. Á¦¾ÈµÈ ¹æ½ÄÀº ÇнÀ¿¡ »ç¿ëµÇ´Â ¹®ÀåÀÇ ´Ü¾îµé Áß ÀϺθ¦ °íÀ¯¸í»ç »çÀüÀ» ÀÌ¿ëÇÏ¿© ±âÈ£ÈÇÏ°í, ±âÈ£ÈµÈ ´Ü¾îµéÀ» Æ÷ÇÔÇÑ ¹®Àåµé·Î ¹ø¿ª ¸ðµ¨À» ÇнÀ½ÃŲ´Ù. »õ·Î¿î ¹®Àå ¹ø¿ª½Ã¿¡µµ °íÀ¯¸í»ç »çÀüÀ» ÀÌ¿ëÇÏ¿© ±âÈ£ÈÇÏ°í ¹ø¿ªÈÄ º¹È£È ÇÏ´Â ¹æ½ÄÀ¸·Î ¹ø¿ªÀ» ¿Ï¼ºÇÑ´Ù. Á¦¾ÈµÈ ¹æ½ÄÀÇ ¼º´ÉÀ» °ËÁõÇϱâ À§ÇØ °íÀ¯¸í»ç ±âȣȸ¦ »ç¿ëÇÏÁö ¾ÊÀº ¸ðµ¨°ú ÇÔ²² ºñ±³ ½ÇÇèÇÏ¿´°í, BLEU Á¡¼ö¸¦ ÅëÇØ ¼öÄ¡ÀûÀ¸·Î °³¼±µÇ´Â °æ¿ìµéµµ È®ÀÎÇßÀ¸¸ç, ¸î°¡Áö ¹ø¿ª »ç·Êµéµµ »ó¿ë¼ºñ½º °á°úµé°ú ÇÔ²² Á¦½ÃÇß´Ù. |
¿µ¹®³»¿ë (English Abstract) |
There is progress in the field of neural machine translation, but there are cases where the translation of sentences containing proper nouns, such as, names, new words, and words that are used only within a specific group, is not accurate. To handle such cases, this paper uses the Korean-English proper noun dictionary and the symbolization method in addition to the recently proposed translation model, Transformer Model. In the proposed method, some of the words in the sentences used for learning are symbolized using a proper noun dictionary, and the translation model is trained with sentences including the symbolized words. When translating a new sentence, the translation is completed by symbolizing, translation, and desymbolizing. The proposed method was compared with a model without symbolization, and for some cases improvement was quantitatively confirmed with the BLEU score. In addition, several examples of translation were also presented along with commercial service results. |
Å°¿öµå(Keyword) |
½Å°æ ±â°è ¹ø¿ª
°íÀ¯¸í»ç ¹ø¿ª
±âÈ£È
°íÀ¯¸í»ç »çÀü
neural machine translation
proper noun translation
symbolization
proper noun dictionary
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|