Á¤º¸Ã³¸®ÇÐȸ ³í¹®Áö ¼ÒÇÁÆ®¿þ¾î ¹× µ¥ÀÌÅÍ °øÇÐ
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
Çѱ¹¾î ¾ð¾î ¸ðµ¨À» È°¿ëÇÑ º¸À̽ºÇÇ½Ì Å½Áö ±â´É °³¼± |
¿µ¹®Á¦¸ñ(English Title) |
Exploiting Korean Language Model to Improve Korean Voice Phishing Detection |
ÀúÀÚ(Author) |
Milandu Keith Moussavou Boussougou
¹Úµ¿ÁÖ
Milandu Keith Moussavou Boussougou
Dong-Joo Park
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 11 NO. 10 PP. 0437 ~ 0446 (2022. 10) |
Çѱ۳»¿ë (Korean Abstract) |
º¸À̽ºÇÇ½Ì ÅëÈ ³»¿ëÀ» ŽÁöÇÏ°í ºÐ·ùÇϴµ¥ ÇÙ½É ¿£ÁøÀ¸·Î ÃֽŠ¸Ó½Å·¯´×(ML) ¹× µö·¯´×(DL) ¾Ë°í¸®Áò°ú °áÇÕµÈ ÀÚ¿¬¾î ó¸®(NLP)ÀÇ ÅؽºÆ® ºÐ·ù ÀÛ¾÷ÀÌ ³Î¸® »ç¿ëµÈ´Ù. ºñ´ë¸é ±ÝÀ¶°Å·¡ÀÇ Áõ°¡¿Í ´õºÒ¾î º¸À̽ºÇÇ½Ì ÅëÈ ³»¿ë ºÐ·ù¿¡ ´ëÇÑ ¸¹Àº ¿¬±¸°¡ ÁøÇàµÇ°í ¾çÈ£ÇÑ ¼º°ú¸¦ º¸ÀÌ°í ÀÖÁö¸¸, ÃֽŠNLP ±â¼úÀ» È°¿ëÇÑ ¼º´É °³¼±ÀÇ Çʿ伺ÀÌ ¿©ÀüÈ÷ Á¸ÀçÇÑ´Ù. º» ³í¹®Àº KorCCVi¶ó´Â ·¹À̺íÀÌ ÁöÁ¤µÈ Çѱ¹ º¸À̽º ÇÇ½Ì µ¥ÀÌÅÍÀÇ ÅؽºÆ® ºÐ·ù¸¦ ±â¹ÝÀ¸·Î ¿©·¯ ´Ù¸¥ ÃֽŠ¾Ë°í¸®Áò°ú ºñ±³ÇÏ¿© »çÀü ÈÆ·ÃµÈ Çѱ¹¾î ¸ðµ¨ KoBERTÀÇ Çѱ¹ º¸À̽º ÇÇ½Ì Å½Áö ¼º´ÉÀ» º¥Ä¡¸¶Å·ÇÑ´Ù. ½ÇÇè °á°ú¿¡ µû¸£¸é KoBERT ¸ðµ¨ÀÇ Å×½ºÆ® ÁýÇÕ¿¡¼ ºÐ·ù Á¤È®µµ°¡ 99.60%·Î ´Ù¸¥ ¸ðµç ¸ðµ¨ÀÇ ¼º´ÉÀ» ´É°¡ÇÑ´Ù. |
¿µ¹®³»¿ë (English Abstract) |
Text classification task from Natural Language Processing (NLP) combined with state-of-the-art (SOTA) Machine Learning (ML) and Deep Learning (DL) algorithms as the core engine is widely used to detect and classify voice phishing call transcripts. While numerous studies on the classification of voice phishing call transcripts are being conducted and demonstrated good performances, with the increase of non-face-to-face financial transactions, there is still the need for improvement using the latest NLP technologies. This paper conducts a benchmarking of Korean voice phishing detection performances of the pre-trained Korean language model KoBERT, against multiple other SOTA algorithms based on the classification of related transcripts from the labeled Korean voice phishing dataset called KorCCVi. The results of the experiments reveal that the classification accuracy on a test set of the KoBERT model outperforms the performances of all other models with an accuracy score of 99.60%. |
Å°¿öµå(Keyword) |
KoBERT
ÀÚ¿¬¾î ó¸®
ÅؽºÆ® ºÐ·ù
¸Ó½Å·¯´×
µö·¯´×
KoBERT
NLP
Text Classification
Machine Learning
Deep Learning
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|