• 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¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) LUKE¸¦ ÀÌ¿ëÇÑ Çѱ¹¾î ÀÚ¿¬¾î ó¸®: °³Ã¼¸í ÀνÄ, °³Ã¼ ¿¬°á
¿µ¹®Á¦¸ñ(English Title) LUKE for Korean Natural Language Processing: Named Entity Recognition and Entity Linking
ÀúÀÚ(Author) ¹ÎÁø¿ì   ³ª½ÂÈÆ   ±èÇöÈ£   ±è¼±ÈÆ   °­ÀÎÈ£   Jinwoo Min   Seung-Hoon Na   Hyun-Ho Kim   Seon-Hoon Kim   Inho Kang  
¿ø¹®¼ö·Ïó(Citation) VOL 28 NO. 03 PP. 0175 ~ 0183 (2022. 03)
Çѱ۳»¿ë
(Korean Abstract)
BERT¿Í °°Àº Æ®·£½ºÆ÷¸Ó ±â¹ÝÀÇ ¾ð¾î ¸ðµ¨Àº ´ë¿ë·®ÀÇ ·¹À̺íÀÌ ¾ø´Â ¸»¹¶Ä¡¸¦ ÀÚ°¡ ÇнÀ ¹æ¹ýÀ» ÅëÇØ ÇнÀÇÑ ÈÄ ´Ù¾çÇÑ ÀÚ¿¬¾î ó¸® ÀÀ¿ë ŽºÅ©¿¡ Àû¿ëÇÏ¿© ³î¶ó¿î ¼º´É Çâ»óÀ» º¸¿´´Ù. ÀÌ¿Í °°Àº ¾ð¾î ¸ðµ¨Àº ½Ç¼¼°è Áö½Ä Á¤º¸¸¦ Ç¥ÇöÇÒ ¼ö ¾ø´Â ´ÜÁ¡ÀÌ Á¸ÀçÇÏ°í ÀÌ·¯ÇÑ ¹®Á¦¸¦ ÇØ°áÇϱâ À§ÇØ ¾ð¾î ¸ðµ¨¿¡ Áö½Ä º£À̽º¸¦ ¹Ý¿µÇÏ·Á´Â ´Ù¾çÇÑ ¿¬±¸µéÀÌ ¼öÇàµÇ¾ú´Ù. º» ¿¬±¸¿¡¼­´Â ´Ü¾î ½ÃÄö½º ÀÌ¿Ü¿¡ ¿£Æ¼Æ¼ ½ÃÄö½º¿Í ÀÓº£µùÀ» Á¤ÀÇÇÏ°í ´Ü¾î¿Í ¿£Æ¼Æ¼ÀÇ ¸ðµç ½ÃÄö½º ½Ö¿¡ µû¶ó º°µµÀÇ Äõ¸® ÆĶó¹ÌÅ͸¦ µÎ°í ¼¿ÇÁ ¾îÅÙ¼ÇÀ» ¼öÇàÇÏ´Â LUKE ¸ðµ¨À» Çѱ¹¾î À§Å°Çǵð¾Æ »ó¿¡¼­ ÇнÀÇÑ ÈÄ ¿£Æ¼Æ¼ °ü·Ã ŽºÅ©ÀÎ °³Ã¼¸í ÀνÄ, °³Ã¼ ¿¬°á¿¡ Àû¿ëÇÏ¿© ±âÁ¸ÀÇ RoBERTa ±â¹Ý ¸ðµ¨ ´ëºñ °¢°¢ 0.5%p, 1.05%pÀÇ ¼º´É Çâ»óÀ» °¡Á®¿Ô´Ù.
¿µ¹®³»¿ë
(English Abstract)
Transformer-based language models (LM) such as BERT trained from a large amount of unlabeled corpus using self-supervised learning methods have shown remarkable performance improvement on various natural language processing (NLP) application tasks. Despite the marked improvements, the classical pretrained language model has not directly incorporate external real-world knowledge bases such as a Wikipedia knowledge graph or triples. To inject the real-world knowledge bases to a pretrained language model, many studies towards ¡°knowledge enhanced¡± pretrained language models have been conducted. Among them, LUKE attaches a sequence of entities to a sequence of original input tokens and performs entity-aware self-attention using entity embeddings, leading to noticeable improved results on entity-related tasks and the state-of-the-art performance in SQuAD dataset. In this paper, we present a Korean version of LUKE pretrained from a large amount of Korean Wikipedia corpus and show its application results on entity-related tasks of Korean. In particular, we newly propose a way of applying LUKE to the entity linking task which has not been explored in the previous works of using LUKE. Experiment results on both Korean named entity recognition and entity linking tasks show improvements over the RoBERTa-based models.
Å°¿öµå(Keyword)
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå