\n1. \uc11c\ub860<\/h2>\n
\n \ucd5c\uadfc \uba87 \ub144\uac04 \uc778\uacf5\uc9c0\ub2a5 \ubd84\uc57c\uc5d0\uc11c \ub525 \ub7ec\ub2dd(deep learning)\uc758 \ubc1c\uc804\uc740 \ub9e4\uc6b0 \ub208\ubd80\uc2e0 \ubcc0\ud654\uc640 \ud601\uc2e0\uc744 \uac00\uc838\uc654\uc2b5\ub2c8\ub2e4. \ub525 \ub7ec\ub2dd\uc740 \ud2b9\ud788 \uc790\uc5f0\uc5b4 \ucc98\ub9ac(Natural Language Processing, NLP) \ubd84\uc57c\uc5d0\uc11c \uc911\uc694\ud55c \uc5ed\ud560\uc744 \ud558\uace0 \uc788\uc73c\uba70, \uadf8\uc5d0 \ub530\ub77c \ub2e4\uc591\ud55c \ubaa8\ub378\ub4e4\uc774 \uac1c\ubc1c\ub418\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774 \uae00\uc5d0\uc11c\ub294 \ud53c\ub4dc \ud3ec\uc6cc\ub4dc \uc2e0\uacbd\ub9dd \uc5b8\uc5b4 \ubaa8\ub378(Neural Network Language Model, NNLM)\uc5d0 \ub300\ud574 \uc54c\uc544\ubcf4\uace0, \uc774 \ubaa8\ub378\uc774 \uc790\uc5f0\uc5b4 \ucc98\ub9ac \ubd84\uc57c\uc5d0\uc11c \uc5b4\ub5bb\uac8c \ud65c\uc6a9\ub420 \uc218 \uc788\ub294\uc9c0, \uadf8\ub9ac\uace0 \uadf8 \uc131\ub2a5\uc744 \ud5a5\uc0c1\uc2dc\ud0a4\uae30 \uc704\ud55c \ub2e4\uc591\ud55c \uae30\ubc95\ub4e4\uc744 \ud0d0\uad6c\ud574 \ubcf4\uaca0\uc2b5\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n2. \uc790\uc5f0\uc5b4 \ucc98\ub9ac(NLP)\uc758 \uac1c\ub150<\/h2>\n
\n \uc790\uc5f0\uc5b4 \ucc98\ub9ac\ub780, \ucef4\ud4e8\ud130\uc640 \uc778\uac04\uc758 \uc5b8\uc5b4 \uac04\uc758 \uc0c1\ud638\uc791\uc6a9\uc744 \ub2e4\ub8e8\ub294 \uc778\uacf5\uc9c0\ub2a5 \ubd84\uc57c\uc785\ub2c8\ub2e4. \uc774 \ubd84\uc57c\ub294 \ud14d\uc2a4\ud2b8\ub098 \uc74c\uc131\uc744 \uc774\ud574\ud558\uace0 \ucc98\ub9ac\ud558\ub294 \uac83\uc744 \ubaa9\ud45c\ub85c \ud558\uba70, \ub2e4\uc591\ud55c \uc560\ud50c\ub9ac\ucf00\uc774\uc158\uc774 \uc874\uc7ac\ud569\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uba74 \uae30\uacc4 \ubc88\uc5ed, \uac10\uc815 \ubd84\uc11d, \uc815\ubcf4 \uac80\uc0c9 \ub4f1\uc774 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ub7ec\ud55c \uc790\uc5f0\uc5b4 \ucc98\ub9ac\uc758 \uae30\ucd08\uac00 \ub418\ub294 \ud575\uc2ec \uae30\uc220 \uc911 \ud558\ub098\uac00 \ubc14\ub85c \uc5b8\uc5b4 \ubaa8\ub378(language model)\uc785\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n3. \uc5b8\uc5b4 \ubaa8\ub378\uc758 \uc815\uc758<\/h2>\n
\n \uc5b8\uc5b4 \ubaa8\ub378\uc740 \ud2b9\uc815 \uc5b8\uc5b4\uc758 \ud1b5\uacc4\uc801 \ud2b9\uc131\uc744 \ud65c\uc6a9\ud558\uc5ec \uc8fc\uc5b4\uc9c4 \uc2dc\ud000\uc2a4\uc758 \ub2e4\uc74c \ub2e8\uc5b4\ub97c \uc608\uce21\ud558\ub294 \ubaa8\ub378\uc785\ub2c8\ub2e4. \uc774 \ubaa8\ub378\uc740 \ub2e8\uc5b4\uc758 \ud655\ub960 \ubd84\ud3ec\ub97c \ud559\uc2b5\ud558\uc5ec \ubb38\uc7a5\uc774 \uc790\uc5f0\uc2a4\ub7fd\uace0 \uc758\ubbf8 \uc788\uac8c \uc0dd\uc131\ub420 \uc218 \uc788\ub3c4\ub85d \ud569\ub2c8\ub2e4. \uc5b8\uc5b4 \ubaa8\ub378\uc758 \ubaa9\uc801\uc740 \ubb38\ubc95\uc801\uc774\uba70, \uc758\ubbf8\ub860\uc801\uc73c\ub85c \uc62c\ubc14\ub978 \ubb38\uc7a5\uc744 \uc0dd\uc131\ud558\ub294 \uac83\uc785\ub2c8\ub2e4.\n <\/p>\n
3.1 \uc804\ud1b5\uc801\uc778 \uc5b8\uc5b4 \ubaa8\ub378<\/h3>\n
\n \uc804\ud1b5\uc801\uc778 \uc5b8\uc5b4 \ubaa8\ub378\uc5d0\ub294 n-\uadf8\ub7a8 \ubaa8\ub378\uacfc \uac19\uc740 \ud1b5\uacc4\uc801 \uc811\uadfc \ubc29\uc2dd\uc774 \ud3ec\ud568\ub429\ub2c8\ub2e4. n-\uadf8\ub7a8 \ubaa8\ub378\uc740 \ub2e8\uc5b4\uc758 \uc5f0\uc18d\ub41c n\uac1c\uc758 \uc2dc\ud000\uc2a4\ub97c \ud1b5\ud574 \ub2e4\uc74c \ub2e8\uc5b4\uc758 \ud655\ub960\uc744 \uacc4\uc0b0\ud569\ub2c8\ub2e4. \uadf8\ub7ec\ub098 \uc774 \ubc29\ubc95\uc740 \uba54\ubaa8\ub9ac\uac00 \ub9ce\uc774 \ud544\uc694\ud558\uace0, \ub370\uc774\ud130\uac00 \ud76c\uc18c\ud560 \ub54c \uc131\ub2a5\uc774 \ub5a8\uc5b4\uc9c0\ub294 \ub2e8\uc810\uc774 \uc788\uc2b5\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n4. \ub525 \ub7ec\ub2dd\uc758 \ub3c4\uc785<\/h2>\n
\n \ucd5c\uadfc\uc5d0\ub294 \ub525 \ub7ec\ub2dd \uae30\ubc95\uc774 \uc804\ud1b5\uc801\uc778 \uc5b8\uc5b4 \ubaa8\ub378\uc744 \ub300\uccb4\ud558\uace0 \uc788\uc2b5\ub2c8\ub2e4. \ud2b9\ud788 \uc2e0\uacbd\ub9dd \uae30\ubc18\uc758 \ubaa8\ub378\ub4e4\uc774 \ub192\uc740 \uc131\ub2a5\uc744 \ubcf4\uc5ec\uc8fc\uba74\uc11c \uc8fc\ubaa9\ubc1b\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ub7ec\ud55c \ub525 \ub7ec\ub2dd \ubaa8\ub378\ub4e4\uc740 \ub9ce\uc740 \uc591\uc758 \ub370\uc774\ud130\ub85c\ubd80\ud130 \ubcf5\uc7a1\ud55c \ud328\ud134\uc744 \ud559\uc2b5\ud560 \uc218 \uc788\uc73c\uba70, \uc774\ub97c \ud1b5\ud574 \ub354\uc6b1 \ubc1c\uc804\ub41c \uc790\uc5f0\uc5b4 \ucc98\ub9ac \uae30\ub2a5\uc744 \uc81c\uacf5\ud569\ub2c8\ub2e4.\n <\/p>\n
4.1 \uc2e0\uacbd\ub9dd \uc5b8\uc5b4 \ubaa8\ub378(Neural Network Language Model)<\/h3>\n
\n \uc2e0\uacbd\ub9dd \uc5b8\uc5b4 \ubaa8\ub378(NNLM)\uc740 \uba3c\uc800 \uc8fc\uc5b4\uc9c4 \ub2e8\uc5b4 \uc2dc\ud000\uc2a4\ub97c \uc785\ub825\ubc1b\uc544 \uac01\uac01\uc758 \ub2e8\uc5b4\ub97c \ubca1\ud130\ub85c \ubcc0\ud658\ud569\ub2c8\ub2e4. \uc774\ud6c4 \ud559\uc2b5\ub41c \uc2e0\uacbd\ub9dd\uc744 \ud1b5\ud574 \ub2e4\uc74c \ub2e8\uc5b4\uc758 \ud655\ub960\uc744 \uc608\uce21\ud558\ub294 \uacfc\uc815\uc744 \uac70\uce69\ub2c8\ub2e4. \uc774 \ubaa8\ub378\uc740 \uc804\ud1b5\uc801\uc778 n-\uadf8\ub7a8 \ubaa8\ub378\uc5d0 \ube44\ud574 \ub9ce\uc740 \uc7a5\uc810\uc744 \uac00\uc9c0\uace0 \uc788\uc73c\uba70, \ud2b9\ud788 \ub354 \uae34 \uc758\uc874 \uad00\uacc4\ub97c \ud559\uc2b5\ud560 \uc218 \uc788\ub2e4\ub294 \uc810\uc5d0\uc11c \uc6b0\uc218\ud55c \uc131\ub2a5\uc744 \ubcf4\uc5ec\uc90d\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n5. NNLM\uc758 \uad6c\uc870<\/h2>\n
\n NNLM\uc758 \uad6c\uc870\ub294 \uae30\ubcf8\uc801\uc73c\ub85c \uc785\ub825\uce35, \uc740\ub2c9\uce35, \ucd9c\ub825\uce35\uc758 \uc138 \ubd80\ubd84\uc73c\ub85c \ub098\ub20c \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc785\ub825\uce35\uc5d0\uc11c\ub294 \ub2e8\uc5b4 \ubca1\ud130\ub97c \ubc1b\uc544\ub4e4\uc774\uace0, \uc740\ub2c9\uce35\uc5d0\uc11c\ub294 \uc774 \ubca1\ud130\ub4e4\uc744 \uae30\ubc18\uc73c\ub85c \uc5ec\ub7ec \uac1c\uc758 \ub274\ub7f0\uc774 \ud65c\uc131\ud654\ub429\ub2c8\ub2e4. \ub9c8\uc9c0\ub9c9\uc73c\ub85c \ucd9c\ub825\uce35\uc5d0\uc11c\ub294 \uc608\uce21\ub41c \ub2e8\uc5b4\uc758 \ud655\ub960 \ubd84\ud3ec\ub97c \uc0dd\uc131\ud569\ub2c8\ub2e4.\n <\/p>\n
5.1 \uc785\ub825\uce35<\/h3>\n
\n \uc785\ub825\uce35\uc5d0\uc11c\ub294 \ub2e8\uc5b4\ub97c \uace0\uc815\ub41c \ud06c\uae30\uc758 \ubca1\ud130\ub85c \ubcc0\ud658\ud558\uae30 \uc704\ud574 \uc784\ubca0\ub529(embedding) \uae30\ubc95\uc744 \uc0ac\uc6a9\ud569\ub2c8\ub2e4. \uc774 \uacfc\uc815\uc5d0\uc11c \uac01 \ub2e8\uc5b4\ub294 \uace0\uc720\ud55c \uc2e4\uc218 \ubca1\ud130\ub85c \ud45c\ud604\ub418\uba70, \ubaa8\ub378\uc740 \uc774 \ubca1\ud130\ub4e4\uc744 \uc785\ub825\uc73c\ub85c \ubc1b\uc544\ub4e4\uc785\ub2c8\ub2e4.\n <\/p>\n
5.2 \uc740\ub2c9\uce35<\/h3>\n
\n \uc740\ub2c9\uce35\uc740 \uc5ec\ub7ec \uac1c\uc758 \ub274\ub7f0\uc73c\ub85c \uad6c\uc131\ub418\uba70, \uc785\ub825\ub41c \ub2e8\uc5b4 \ubca1\ud130\uc640 \uac00\uc911\uce58\ub97c \uacf1\ud558\uc5ec \ud65c\uc131\ud654 \ud568\uc218\ub97c \ud1b5\uacfc\uc2dc\ud0b5\ub2c8\ub2e4. \uc77c\ubc18\uc801\uc73c\ub85c \uc0ac\uc6a9\ub418\ub294 \ud65c\uc131\ud654 \ud568\uc218\ub85c\ub294 \ube44\uc120\ud615\uc131\uc744 \ubd80\uc5ec\ud558\uae30 \uc704\ud574 ReLU(Rectified Linear Unit) \ub610\ub294 \uc2dc\uadf8\ubaa8\uc774\ub4dc \ud568\uc218\uc640 \uac19\uc740 \ud568\uc218\uac00 \uc0ac\uc6a9\ub429\ub2c8\ub2e4.\n <\/p>\n
5.3 \ucd9c\ub825\uce35<\/h3>\n
\n \ucd9c\ub825\uce35\uc5d0\uc11c\ub294 softmax \ud568\uc218\ub97c \uc0ac\uc6a9\ud558\uc5ec \uac01 \ub2e8\uc5b4\uc758 \uc608\uce21 \ud655\ub960\uc744 \uacc4\uc0b0\ud569\ub2c8\ub2e4. softmax \ud568\uc218\ub294 \ubaa8\ub4e0 \ub2e8\uc5b4\uc758 \ud655\ub960 \ud569\uc774 1\uc774 \ub418\ub3c4\ub85d \uc815\uaddc\ud654\ud558\uc5ec, \uac00\uc7a5 \ub192\uc740 \ud655\ub960\uc744 \uac00\uc9c4 \ub2e8\uc5b4\ub97c \uc120\ud0dd\ud560 \uc218 \uc788\ub3c4\ub85d \ud569\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n6. NNLM\uc758 \ud559\uc2b5 \uacfc\uc815<\/h2>\n
\n NNLM\uc740 \uc77c\ubc18\uc801\uc778 \uc2e0\uacbd\ub9dd\uacfc \uc720\uc0ac\ud55c \ud559\uc2b5 \uacfc\uc815\uc744 \ub530\ub985\ub2c8\ub2e4. \ub370\uc774\ud130\uc14b\uc744 \ud1b5\ud574 \ubaa8\ub378\uc758 \uac00\uc911\uce58\ub97c \uc5c5\ub370\uc774\ud2b8\ud558\uba70 \uc190\uc2e4 \ud568\uc218\ub294 \uad50\ucc28 \uc5d4\ud2b8\ub85c\ud53c \uc190\uc2e4(cross-entropy loss)\uc774 \uc77c\ubc18\uc801\uc73c\ub85c \uc0ac\uc6a9\ub429\ub2c8\ub2e4.\n <\/p>\n
6.1 \ub370\uc774\ud130 \uc804\ucc98\ub9ac<\/h3>\n
\n \ub370\uc774\ud130 \uc804\ucc98\ub9ac\ub294 \uc2e0\uacbd\ub9dd \uc5b8\uc5b4 \ubaa8\ub378\uc758 \uc131\ub2a5\uc744 \uc88c\uc6b0\ud558\ub294 \uc911\uc694\ud55c \uacfc\uc815\uc785\ub2c8\ub2e4. \ub2e8\uc5b4\ub97c \ubca1\ud130\ub85c \uc784\ubca0\ub529\ud558\uae30 \uc704\ud574, \ud14d\uc2a4\ud2b8 \ub370\uc774\ud130\ub97c \ud1a0\ud070\ud654(tokenization)\ud558\uace0, \ubd88\uc6a9\uc5b4(stopword)\ub97c \uc81c\uac70\ud558\uba70, \ub2e8\uc5b4\uc758 \ube48\ub3c4\uc218\ub97c \uae30\ubc18\uc73c\ub85c \uc801\uc808\ud55c \ub2e8\uc5b4 \uc9d1\ud569(vocabulary)\uc744 \uc0dd\uc131\ud558\ub294 \ub4f1\uc758 \uc791\uc5c5\uc774 \ud544\uc694\ud569\ub2c8\ub2e4.\n <\/p>\n
6.2 \uc190\uc2e4 \ud568\uc218 \ubc0f \ucd5c\uc801\ud654<\/h3>\n
\n NNLM\uc758 \uc190\uc2e4 \ud568\uc218\ub294 \uc608\uce21 \ud655\ub960\uacfc \uc2e4\uc81c \ub2e8\uc5b4 \uac04\uc758 \ucc28\uc774\ub97c \uacc4\uc0b0\ud569\ub2c8\ub2e4. \uc774\ub97c \ud1b5\ud574 Backpropagation\uc744 \uc0ac\uc6a9\ud558\uc5ec \uac00\uc911\uce58\ub97c \uc5c5\ub370\uc774\ud2b8\ud558\uace0 \ubaa8\ub378\uc744 \ud559\uc2b5\uc2dc\ud0b5\ub2c8\ub2e4. \ucd5c\uc801\ud654 \uc54c\uace0\ub9ac\uc998\uc73c\ub85c\ub294 SGD(Stochastic Gradient Descent)\ub098 Adam optimizer\uc640 \uac19\uc740 \ubc29\ubc95\uc774 \uc0ac\uc6a9\ub429\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n7. NNLM\uc758 \uc7a5\uc810\uacfc \ud55c\uacc4<\/h2>\n7.1 \uc7a5\uc810<\/h3>\n
\n NNLM\uc758 \uac00\uc7a5 \ud070 \uc7a5\uc810\uc740 \ub2e8\uc5b4 \uac04\uc758 \ubcf5\uc7a1\ud55c \uad00\uacc4\ub97c \ud559\uc2b5\ud560 \uc218 \uc788\ub2e4\ub294 \uac83\uc785\ub2c8\ub2e4. \uc804\ud1b5\uc801\uc778 n-\uadf8\ub7a8 \ubaa8\ub378\uc740 \ud55c\uc815\ub41c \uacfc\uac70\uc758 \ub370\uc774\ud130\ub9cc\uc744 \uace0\ub824\ud558\uc9c0\ub9cc, NNLM\uc740 \ubb38\ub9e5\uc744 \uae30\ubc18\uc73c\ub85c \ud55c \uae34 \uc758\uc874\uc131\uc744 \ud559\uc2b5\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ub294 \uc790\uc5f0\uc5b4 \ucc98\ub9ac\uc5d0\uc11c \ub354 \uc758\ubbf8 \uc788\ub294 \ubb38\uc7a5\uc744 \uc0dd\uc131\ud558\uace0 \uc774\ud574\ud558\ub294 \ub370 \ud070 \ub3c4\uc6c0\uc774 \ub429\ub2c8\ub2e4.\n <\/p>\n
7.2 \ud55c\uacc4<\/h3>\n
\n \ud55c\ud3b8, NNLM\ub3c4 \uba87 \uac00\uc9c0 \ud55c\uacc4\ub97c \uac00\uc9d1\ub2c8\ub2e4. \ub300\ud45c\uc801\uc73c\ub85c\ub294 \ub300\ub7c9\uc758 \ub370\uc774\ud130\uc640 \ucef4\ud4e8\ud305 \uc790\uc6d0\uc744 \uc694\uad6c\ud558\uba70, \ucda9\ubd84\ud55c \ub370\uc774\ud130\uac00 \uc5c6\uc744 \uacbd\uc6b0 \uc131\ub2a5\uc774 \ud06c\uac8c \uc800\ud558\ub420 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ub610\ud55c, \ub2e8\uc5b4\uc758 \uc21c\uc11c\ub098 \ubb38\ub9e5\uc5d0 \ub530\ub77c \uadf8 \uc758\ubbf8\uac00 \ub2e4\ub974\uac8c \ud574\uc11d\ub420 \uc218 \uc788\ub294 \uacbd\uc6b0, \uc774\ub294 \ubaa8\ub378 \ub2a5\ub825\uc744 \uc81c\ud55c\ud558\ub294 \uc694\uc18c\uac00 \ub420 \uc218 \uc788\uc2b5\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n8. NNLM\uc758 \ubc1c\uc804\uacfc \ub2e4\uc591\uc131<\/h2>\n
\n NNLM\uc740 \uae30\ubcf8\uc801\uc778 \uc5b8\uc5b4 \ubaa8\ub378\uc5d0\uc11c \ucd9c\ubc1c\ud588\uc9c0\ub9cc, \ub2e4\uc591\ud55c \ubcc0\uc6a9 \ubaa8\ub378\ub4e4\uc774 \uac1c\ubc1c\ub418\uc5c8\uc2b5\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4, LSTM(Long Short-Term Memory)\uacfc GRU(Gated Recurrent Unit)\uc640 \uac19\uc740 \uc21c\ud658 \uc2e0\uacbd\ub9dd(RNN) \uae30\ubc18\uc758 \ubaa8\ub378\uc740 \uc2dc\uac04\uc744 \uace0\ub824\ud55c \ubb38\ub9e5 \uc815\ubcf4\ub97c \ub354\uc6b1 \ud6a8\uacfc\uc801\uc73c\ub85c \ucea1\ucc98\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ub354\ubd88\uc5b4 Transformer \ubaa8\ub378\uc740 Attention \uba54\ucee4\ub2c8\uc998\uc744 \uc774\uc6a9\ud558\uc5ec \uc7a5\uae30\uc801\uc778 \uc758\uc874 \uad00\uacc4\ub97c \ub354\uc6b1 \uc798 \ubaa8\ub378\ub9c1\ud558\ub294 \ub370 \uae30\uc5ec\ud558\uace0 \uc788\uc2b5\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n9. \uc2e4\ud5d8\uacfc \ud3c9\uac00<\/h2>\n
\n NNLM\uc758 \uc131\ub2a5\uc744 \ud3c9\uac00\ud558\uae30 \uc704\ud574 \ub2e4\uc591\ud55c \ub370\uc774\ud130\uc14b\uacfc \ud3c9\uac00 \uc9c0\ud45c\uac00 \uc0ac\uc6a9\ub429\ub2c8\ub2e4. \ub300\ud45c\uc801\uc778 \ub370\uc774\ud130\uc14b\uc73c\ub85c\ub294 Penn Treebank, WikiText \ub4f1\uc774 \uc788\uc73c\uba70, PERPLEXITY \ub610\ub294 \uc815\ud655\ub3c4(accuracy), F1 \uc810\uc218(F1 score)\uc640 \uac19\uc740 \ud3c9\uac00 \uc9c0\ud45c\uac00 \ud65c\uc6a9\ub429\ub2c8\ub2e4.\n <\/p>\n<\/section>\n\n10. \uacb0\ub860<\/h2>\n
\n \ud53c\ub4dc \ud3ec\uc6cc\ub4dc \uc2e0\uacbd\ub9dd \uc5b8\uc5b4 \ubaa8\ub378(Neural Network Language Model, NNLM)\uc740 \ub525 \ub7ec\ub2dd\uc758 \ubc1c\uc804\uacfc \ud568\uaed8 \uc790\uc5f0\uc5b4 \ucc98\ub9ac \ubd84\uc57c\uc5d0\uc11c \uc911\uc694\ud55c \uc5ed\ud560\uc744 \ucc28\uc9c0\ud558\uace0 \uc788\uc2b5\ub2c8\ub2e4. \ubcf8 \uae00\uc5d0\uc11c\ub294 NNLM\uc758 \uc774\ub860\uc801 \ubc30\uacbd\uacfc \uad6c\uc870, \ud559\uc2b5 \uacfc\uc815, \uc7a5\ub2e8\uc810 \ub4f1\uc744 \uc0b4\ud3b4\ubcf4\uc558\uc2b5\ub2c8\ub2e4. AI\uc640 NLP\uc758 \ubbf8\ub798\ub294 \uc6b0\ub9ac\uac00 \uc54c\uace0 \uc788\ub294 \uc5b8\uc5b4 \ubaa8\ub378\uc744 \uae30\ubc18\uc73c\ub85c \ub354\uc6b1 \ubc1c\uc804\ud560 \uac83\uc774\uba70, NNLM\uacfc \uadf8 \ubcc0\ud615 \ubaa8\ub378\ub4e4\uc740 \uc55e\uc73c\ub85c\ub3c4 \ub9ce\uc740 \uc5f0\uad6c\uc640 \ubc1c\uc804\uc744 \uac70\ub4ed\ud560 \uac83\uc785\ub2c8\ub2e4.\n <\/p>\n<\/section>\n