• 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 : 66 / 69

ÇѱÛÁ¦¸ñ(Korean Title) ¾ð¾î Á¤º¸°¡ ¹Ý¿µµÈ ¹®Àå Á¡¼ö¸¦ È°¿ëÇÏ´Â »èÁ¦ ±â¹Ý ¹®Àå ¾ÐÃà
¿µ¹®Á¦¸ñ(English Title) Deletion-Based Sentence Compression Using Sentence Scoring Reflecting Linguistic Information
ÀúÀÚ(Author) ÀÌÁعü   ±è¼Ò¾ð   ¹Ú¼º¹è   ÀÌÁعü   ±è¼Ò¾ð   ¹Ú¼º¹è  
¿ø¹®¼ö·Ïó(Citation) VOL 11 NO. 03 PP. 0125 ~ 0132 (2022. 03)
Çѱ۳»¿ë
(Korean Abstract)
¹®Àå ¾ÐÃàÀº ¿øº» ¹®ÀåÀÇ Áß¿äÇÑ Àǹ̴ À¯ÁöÇϸ鼭 ±æÀÌ°¡ Ãà¼ÒµÈ ¾ÐÃà ¹®ÀåÀ» »ý¼ºÇÏ´Â ÀÚ¿¬¾îó¸® Å½ºÅ©ÀÌ´Ù. ¹®¹ýÀûÀ¸·Î ÀûÀýÇÑ ¹®Àå ¾ÐÃàÀ» À§ÇØ, Ãʱ⠿¬±¸µéÀº »ç¶÷ÀÌ Á¤ÀÇÇÑ ¾ð¾î ±ÔÄ¢À» È°¿ëÇÏ¿´´Ù. ¶ÇÇÑ ½ÃÄö½º-Åõ-½ÃÄö½º ¸ðµ¨ÀÌ ±â°è ¹ø¿ª°ú °°Àº ´Ù¾çÇÑ ÀÚ¿¬¾îó¸® Å½ºÅ©¿¡¼­ ÁÁÀº ¼º´ÉÀ» º¸À̸鼭, À̸¦ ¹®Àå ¾ÐÃà¿¡ È°¿ëÇÏ°íÀÚ ÇÏ´Â ¿¬±¸µéµµ Á¸ÀçÇß´Ù. ÇÏÁö¸¸ ¾ð¾î ±ÔÄ¢À» È°¿ëÇÏ´Â ¿¬±¸ÀÇ °æ¿ì ¸ðµç ¾ð¾î ±ÔÄ¢À» Á¤ÀÇÇÏ´Â µ¥¿¡ Å« ºñ¿ëÀÌ µé°í, ½ÃÄö½º-Åõ-½ÃÄö½º ¸ðµ¨ ±â¹Ý ¿¬±¸ÀÇ °æ¿ì ÇнÀÀ» À§ÇØ ´ë·®ÀÇ µ¥ÀÌÅͼÂÀÌ ÇÊ¿äÇÏ´Ù´Â ¹®Á¦Á¡ÀÌ Á¸ÀçÇÑ´Ù. À̸¦ ÇØ°áÇÒ ¼ö ÀÖ´Â ¹æ¹ýÀ¸·Î »çÀü ÇнÀµÈ ¾ð¾î ¸ðµ¨ÀÎ BERT¸¦ È°¿ëÇÏ´Â ¹®Àå ¾ÐÃà ¸ðµ¨ÀÎ Deleter°¡ Á¦¾ÈµÇ¾ú´Ù. Deleter´Â BERT¸¦ ÅëÇØ °è»êµÈ perplexity¸¦ È°¿ëÇÏ¿© ¹®ÀåÀ» ¾ÐÃàÇϱ⠶§¹®¿¡ ¹®Àå ¾ÐÃà ±ÔÄ¢°ú ¸ðµ¨ ÇнÀÀ» À§ÇÑ µ¥ÀÌÅͼÂÀÌ ÇÊ¿äÇÏÁö ¾Ê´Ù´Â ÀåÁ¡ÀÌ ÀÖ´Ù. ÇÏÁö¸¸ Deleter´Â perplexity¸¸À» °í·ÁÇÏ¿© ¹®ÀåÀ» ¾ÐÃàÇϱ⠶§¹®¿¡, ¹®Àå¿¡ ¼ÓÇÑ ´Ü¾îµéÀÇ ¾ð¾î Á¤º¸¸¦ ¹Ý¿µÇÏ¿© ¹®ÀåÀ» ¾ÐÃàÇÏÁö ¸øÇÑ´Ù. ¶ÇÇÑ, perplexity ÃøÁ¤À» À§ÇÑ BERTÀÇ »çÀü ÇнÀ¿¡ »ç¿ëµÈ µ¥ÀÌÅÍ°¡ ¾ÐÃà ¹®Àå°ú °Å¸®°¡ ÀÖ¾î, À̸¦ ÅëÇØ ÃøÁ¤µÈ perplexity°¡ À߸øµÈ ¹®Àå ¾ÐÃàÀ» À¯µµÇÒ ¼ö ÀÖ´Ù´Â ¹®Á¦Á¡ÀÌ ÀÖ´Ù. À̸¦ ÇØ°áÇϱâ À§ÇØ º» ³í¹®Àº ¾ð¾î Á¤º¸ÀÇ Áß¿äµµ¸¦ ¼öÄ¡È­ÇÏ¿© perplexity ±â¹ÝÀÇ ¹®Àå Á¡¼ö °è»ê¿¡ ¹Ý¿µÇÏ´Â ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. ¶ÇÇÑ °íÀ¯¸í»ç°¡ ÀÚÁÖ Æ÷ÇԵǾî ÀÖÀ¸¸ç, ºÒÇÊ¿äÇÑ ¼ö½Ä¾î°¡ »ý·«µÇ´Â °æ¿ì°¡ ¸¹Àº ´º½º ±â»ç ¸»¹¶Ä¡·Î BERT¸¦ fine-tuningÇÏ¿© ¹®Àå ¾ÐÃà¿¡ ÀûÀýÇÑ perplexity¸¦ ÃøÁ¤ÇÒ ¼ö ÀÖµµ·Ï ÇÏ¿´´Ù. ¿µ¾î ¹× Çѱ¹¾î µ¥ÀÌÅÍ¿¡ ´ëÇÑ ¼º´É Æò°¡¸¦ À§ÇØ º» ³í¹®¿¡¼­ Á¦¾ÈÇÏ´Â LI-Deleter¿Í ºñ±³ ¸ðµ¨ÀÇ ¹®Àå ¾ÐÃà ¼º´ÉÀ» ºñ±³ ½ÇÇèÀ» ÁøÇàÇÏ¿´°í, ³ôÀº ¹®Àå ¾ÐÃà ¼º´ÉÀ» º¸ÀÓÀ» È®ÀÎÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
Sentence compression is a natural language processing task that generates concise sentences that preserves the important meaning of the original sentence. For grammatically appropriate sentence compression, early studies utilized human-defined linguistic rules. Furthermore, while the sequence-to-sequence models perform well on various natural language processing tasks, such as machine translation, there have been studies that utilize it for sentence compression. However, for the linguistic rule-based studies, all rules have to be defined by human, and for the sequence-to-sequence model based studies require a large amount of parallel data for model training. In order to address these challenges, Deleter, a sentence compression model that leverages a pre-trained language model BERT, is proposed. Because the Deleter utilizes perplexity based score computed over BERT to compress sentences, any linguistic rules and parallel dataset is not required for sentence compression. However, because Deleter compresses sentences only considering perplexity, it does not compress sentences by reflecting the linguistic information of the words in the sentences. Furthermore, since the dataset used for pre-learning BERT are far from compressed sentences, there is a problem that this can lad to incorrect sentence compression. In order to address these problems, this paper proposes a method to quantify the importance of linguistic information and reflect it in perplexity-based sentence scoring. Furthermore, by fine-tuning BERT with a corpus of news articles that often contain proper nouns and often omit the unnecessary modifiers, we allow BERT to measure the perplexity appropriate for sentence compression. The evaluations on the English and Korean dataset confirm that the sentence compression performance of sentence-scoring based models can be improved by utilizing the proposed method.
Å°¿öµå(Keyword) ¹®Àå ¾ÐÃà   ¾ð¾î Á¤º¸   ¾ð¾î ¸ðµ¨   ÆÞÇ÷º½ÃƼ   Sentence Compression   Linguistic Information   Language Model   Perplexity  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå