Ç¥ÁØÈ­ Âü¿©¾È³»

TTAÀÇ Ç¥ÁØÇöȲ

Ȩ > Ç¥ÁØÈ­ °³¿ä > TTAÀÇ Ç¥ÁØÇöȲ

Ç¥ÁعøÈ£ TTAK.KO-10.0946 ±¸Ç¥ÁعøÈ£
Á¦°³Á¤ÀÏ 2016-12-27 ÃÑÆäÀÌÁö 253
ÇѱÛÇ¥Áظí ÀÚ¿¬¾î ÁúÀÇÀÀ´äÀ» À§ÇÑ ´ÙÀǾî ű븻¹¶Ä¡ ±¸Ãà Áöħ
¿µ¹®Ç¥Áظí Polysemy Tagged Corpus for Question Answering
Çѱ۳»¿ë¿ä¾à Çѱ¹¾î¿¡¼­ ´Ü¾îÀÇ ÇüÅ´ °°À¸³ª ±×°ÍÀÇ Àǹ̰¡ ´Ù¸¥ °ÍµéÀÌ ÀÖ´Ù. ¿¹¸¦ µé¾î, ¸í»ç ¡®¹è¡¯ÀÇ °æ¿ì, µ¿¹°À̳ª »ç¶÷ÀÇ ±â°üÀ» ÀǹÌÇϱ⵵ ÇÏ°í, ±³Åë¼ö´ÜÀ» ÀǹÌÇϱ⵵ Çϸç, °úÀÏÀ» ÀǹÌÇϱ⵵ ÇÑ´Ù. ÀÌ·¸°Ô ´Ü¾îÀÇ ÇüÅ´ °°Áö¸¸ ¶æÀÌ ´Ù¸¥ ´Ü¾î¸¦ µ¿ÇüÀÌÀǾî(homonymy)¶ó°í ÇÑ´Ù. ±×¸®°í ºñ½ÁÇÑ Àǹ̸¦ °¡Áø ´Ü¾îÀÌÁö¸¸ ¾²ÀÓ¿¡ µû¶ó ±×°ÍÀÇ Àǹ̰¡ ´õ ¼¼ºÐÈ­µÈ °æ¿ìµµ Á¸ÀçÇÑ´Ù. ¿¹¸¦ µé¾î ¡®¹è¡¯´Â »ç¶÷À̳ª µ¿¹°ÀÇ ¸ö¿¡ ÀÖ´Â ¹è¿Í ÀýÁ· µ¿¹°ÀÇ ¹è°¡ Àǹ̻óÀ¸·Î´Â ¸Å¿ì ºñ½ÁÇÏÁö¸¸ ±×°ÍÀÇ À§Ä¡°¡ ´Ù¸£¸ç °¢°¢À» ´Ù¸¥ Àǹ̷ΠºÐ¼®ÇÒ ¼ö ÀÖ´Ù. ÀÌ¿Í °°ÀÌ Æ÷°ýÀûÀÎ Àǹ̴ °°Áö¸¸ ¼¼ºÎÀûÀÎ Àǹ̰¡ ´Ù¸¥ ´Ü¾î¸¦ ´ÙÀǾî(Polysemy)¶ó°í ÇÑ´Ù. ´ÙÀǾî´Â °¢ Àǹ̸¦ ³Ê¹« ¼¼ºÐÈ­Çϱ⠶§¹®¿¡ Àǹ̿¡ µû¶ó¼­´Â ¾Ö¸ÅÇÑ °æ¿ì°¡ ¹ß»ýÇÒ ¼ö ÀÖ´Ù.
ÀÌ Ç¥ÁØÀº ´ÙÀǾî ÅÂ±ë ¸»¹¶Ä¡ ±¸ÃàÀ» À§ÇÑ ÀýÂ÷¿Í űë ÁöħÀ» Á¦½ÃÇϸç, ÀÚ¿¬¾î ÁúÀÇÀÀ´ä¿¡ ÀÚÁÖ µîÀåÇÏ´Â ¿ë¾ðÀ» ´ë»óÀ¸·Î ±¸Ã¼ÀûÀÎ ´ÙÀǾî ÀÇ¹Ì º° ¼¼ºÎÁöħÀ» »ç·Ê·Î ºÎ·Ï¿¡ Á¦°øÇÑ´Ù.
¿µ¹®³»¿ë¿ä¾à In computational linguistics, WSD is identifying which sense of a word is used in a sentence, when the word has multiple meanings. The notion of "word sense" is slippery and controversial. Most people can agree in distinctions at the coarse-grained homograph level (e.g., pen as writing instrument or enclosure), but when we go down one level to fine-grained polysemy, disagreements arise. Word meaning is in principle infinitely variable and context sensitive. It is not divided up easily into distinct or discrete sub-meanings. The standard is to present commonly usable process and guidelines for building a sense tagged corpus and is based on verbs which are frequently used in natural language question answering field.
±¹Á¦Ç¥ÁØ
°ü·ÃÆÄÀÏ TTAK.KO-10.0946.pdf TTAK.KO-10.0946.pdf            

ÀÌÀü
º¹ÇÕ ÀÓº£µðµå ½Ã½ºÅÛ ³»ÀÇ ÀüÀÚÀåÄ¡ °£ µ¥ÀÌÅÍ ±³È¯ ÇÁ·Î±×·¡¹Ö ÀÎÅÍÆäÀ̽º ±â´É ¸í¼¼
´ÙÀ½
½Ç½Ã°£ »çÀ̹ö-¹°¸® ½Ã½ºÅÛ(CPS) ÀÀ¿ëÀ» À§ÇÑ µ¥ÀÌÅͺй輭ºñ½º ¿ä±¸»çÇ×