{"id":1506,"date":"2015-11-27T16:33:27","date_gmt":"2015-11-27T09:33:27","guid":{"rendered":"http:\/\/www.clc.hcmus.edu.vn\/?page_id=1506"},"modified":"2016-06-28T11:33:44","modified_gmt":"2016-06-28T04:33:44","slug":"xay-dung-va-khai-thac-kho-ngu-lieu-song-ngu-anh-viet","status":"publish","type":"page","link":"https:\/\/www.clc.hcmus.edu.vn\/?page_id=1506","title":{"rendered":"X\u00e2y d\u1ef1ng v\u00e0 khai th\u00e1c Kho Ng\u1eef li\u1ec7u Song ng\u1eef Anh-Vi\u1ec7t"},"content":{"rendered":"

 <\/p>\n

\n

X\u00e2y d\u1ef1ng v\u00e0 khai th\u00e1c Kho N<\/span>g\u1eef li\u1ec7u Song ng\u1eef <\/span>Anh-Vi\u1ec7t<\/span>\u00a0(*)<\/span><\/strong><\/p>\n

1.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>T\u1ed4NG QUAN<\/span><\/h1>\n

Trong <\/span>vi\u1ec7c nghi\u00ean c\u1ee9u, g<\/span>i\u1ea3ng d\u1ea1y ng<\/span>\u00f4n<\/span> ng\u1eef, <\/span>ta c\u1ea7n<\/span> th\u1ed1ng k\u00ea, <\/span>s<\/span>o s\u00e1nh<\/span>,<\/span> \u0111\u1ed1i chi\u1ebfu \u0111\u1ec3 t\u00ecm ra c\u00e1c <\/span>quy lu\u1eadt c\u1ee7a ng\u00f4n ng\u1eef, quy lu\u1eadt chuy\u1ec3n ng\u1eef, c\u00e1c <\/span>\u0111i\u1ec3m t\u01b0<\/span>\u01a1ng<\/span> \u0111\u1ed3ng v\u00e0 d\u1ecb bi\u1ec7t \u1edf c\u00e1c b\u00ecnh di\u1ec7n<\/span> kh\u00e1c nh<\/span>au<\/span>, c\u00e1c c\u1ea5p \u0111\u1ed9 kh\u00e1c nhau<\/span> gi\u1eefa c\u00e1c ng\u00f4n ng\u1eef. Nh\u01b0ng \u0111\u1ec3 <\/span>th\u1ed1ng k\u00ea, <\/span>so s\u00e1nh,<\/span> \u0111\u1ed1i chi\u1ebfu nh\u01b0 tr\u00ean,<\/span> ta <\/span>c\u1ea7n <\/span>ph\u1ea3i c\u00f3 c\u1ee9 li\u1ec7u c\u1ee7a c\u00e1c ng\u00f4n ng\u1eef<\/span> m\u00e0 ta \u0111ang c\u1ea7n so s\u00e1nh<\/span>, <\/span>ta g\u1ecdi <\/span>\u0111\u00f3 <\/span>l\u00e0 \u201cng\u1eef li\u1ec7u\u201d<\/span> (<\/span>corpus)<\/span>. <\/span>Ng\u1eef<\/span> li\u1ec7u<\/span> \u1edf \u0111\u00e2y <\/span>\u0111\u01b0\u1ee3c hi\u1ec3u l\u00e0 <\/span>t\u1eadp h\u1ee3p <\/span>v\u0103n b\u1ea3n \u0111\u01a1n ng\u1eef<\/span>,<\/span> \u0111a<\/span> ng\u1eef<\/span> hay song ng\u1eef (g\u1ed3m c\u00e1c c\u1eb7p v\u0103n b\u1ea3n <\/span>\u0111\u00e3<\/span> \u0111\u01b0\u1ee3c d\u1ecbch <\/span>th\u1ee7 c\u00f4ng<\/span>, d\u1ecbch<\/span> t\u01b0\u01a1ng \u1ee9ng 1-1 v\u1ec1 m\u1eb7t ng\u1eef<\/span> ngh\u0129a<\/span>)<\/span> v\u00e0 ph\u00f9 h\u1ee3p v\u1edbi l\u0129nh v\u1ef1c, th\u1ec3 lo\u1ea1i<\/span>, ni\u00ean \u0111\u1ea1i<\/span> m\u00e0 ta c\u1ea7n nghi\u00ean c\u1ee9u<\/span>.<\/span> Ngo\u00e0i p<\/span>h\u1ea7n <\/span>T\u1ed5ng quan v\u00e0 K\u1ebft lu\u1eadn, <\/span>nghi\u00ean c\u1ee9u<\/span> n\u00e0y<\/span> g<\/span>\u1ed3m<\/span> c\u00e1c <\/span>n\u1ed9i dung ch\u00ednh <\/span>sau:<\/span><\/p>\n

–<\/span>\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>Gi\u1edbi thi\u1ec7u k<\/span>ho ng\u1eef li\u1ec7u song ng\u1eef <\/span>Anh \u2013 Vi\u1ec7t<\/span>.<\/span><\/p>\n

–<\/span>\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>X<\/span>\u1eed l\u00fd <\/span>kho <\/span>ng\u1eef li\u1ec7u song ng\u1eef<\/span> Anh \u2013 Vi\u1ec7t<\/span>.<\/span><\/p>\n

–<\/span>\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>K<\/span>hai th\u00e1c <\/span>kho <\/span>ng\u1eef li\u1ec7u song ng\u1eef<\/span> Anh \u2013 Vi\u1ec7t<\/span>.<\/span><\/p>\n

2.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>GI\u1edaI THI\u1ec6U <\/span>KHO <\/span>NG\u1eee LI\u1ec6U SONG NG\u1eee<\/span> ANH \u2013 VI\u1ec6T<\/span><\/h1>\n

Trong <\/span>nghi\u00ean c\u1ee9u<\/span> n\u00e0y, <\/span>ch\u00fang t\u00f4i <\/span>s\u1eed d\u1ee5ng<\/span> ng\u1eef li\u1ec7u <\/span>song song c\u1ee7a 2 ng\u00f4n ng\u1eef <\/span>(<\/span>g\u1ecdi t\u1eaft l\u00e0<\/span> ng\u1eef li\u1ec7u song ng\u1eef<\/span>) <\/span>v\u00e0 c\u1ee5 th\u1ec3 l\u00e0 ng\u1eef li\u1ec7u song ng\u1eef <\/span>gi\u1eefa<\/span> ti\u1ebfng Anh v\u00e0 ti\u1ebfng Vi\u1ec7t <\/span>(<\/span>g\u1ecdi t\u1eaft l\u00e0 <\/span>ng\u1eef li\u1ec7u song ng\u1eef Anh-Vi\u1ec7t<\/span>).<\/span> Trong <\/span>ng\u1eef li\u1ec7u song ng\u1eef<\/span>, c\u00e1c b\u1ea3n d\u1ecbch t\u01b0\u01a1ng \u1ee9ng c\u1ee7a m\u1ed7i ng\u00f4n ng\u1eef ph\u1ea3i \u0111\u01b0\u1ee3c \u0111\u1eb7t song song v\u1edbi nhau hay c\u00f2n \u0111\u01b0\u1ee3c g\u1ecdi l\u00e0 <\/span>gi\u00f3ng h\u00e0ng<\/span> v\u1edbi nhau<\/span> (alignment). <\/span><\/p>\n

\"\"<\/p>\n

H\u00ecnh<\/span> 1. V\u00ed d\u1ee5 gi\u00f3ng h\u00e0ng \u1edf m\u1ee9c \u0111o\u1ea1n.<\/span><\/p>\n

M\u1ee9c \u0111\u1ed9 <\/span>gi\u00f3ng h\u00e0ng<\/span> n\u00e0y c\u00f3 th\u1ec3 \u1edf <\/span>c\u1ea5p \u0111\u1ed9 v\u0103n b\u1ea3n<\/span> (text alignment), <\/span>ngh\u0129a l\u00e0 t\u1eebng v\u0103n b\u1ea3n trong ng\u00f4n ng\u1eef ngu\u1ed3n \u0111\u01b0\u1ee3c gi\u00f3ng (li\u00ean k\u1ebft) v\u1edbi v\u0103n b\u1ea3n d\u1ecbch t\u01b0\u01a1ng \u1ee9ng trong ng\u00f4n ng\u1eef \u0111\u00edch. T\u01b0\u01a1ng t\u1ef1 cho<\/span> c\u1ea5p \u0111\u1ed9 \u0111o\u1ea1n<\/span> (paragraph alignment)<\/span>, <\/span>c\u1ea5p \u0111\u1ed9 c\u00e2u<\/span> (sentence alignment), <\/span>c\u1ea5p \u0111\u1ed9 ng\u1eef<\/span> (phrase alignment) v\u00e0 s\u00e2u nh\u1ea5t l\u00e0 <\/span>c\u1ea5p \u0111\u1ed9 t\u1eeb<\/span> (word alignment). <\/span>H\u00ecnh 1 l\u00e0 m\u1ed9t v\u00ed d\u1ee5 v\u1ec1 gi\u00f3ng h\u00e0ng \u1edf c\u1ea5p \u0111\u1ed9 \u0111o\u1ea1n.<\/span> Trong <\/span>nghi\u00ean c\u1ee9u<\/span> n\u00e0y, ch\u00fang t\u00f4i \u0111i s\u00e2u t\u1edbi <\/span>c\u1ea5p \u0111\u1ed9 <\/span>gi\u00f3ng h\u00e0ng<\/span> t\u1eeb<\/span> (h\u00ecnh 2) <\/span>\u0111\u1ec3 ch\u00fang ta c\u00f3 th\u1ec3 thu \u0111\u01b0\u1ee3c nhi\u1ec1u nh\u1ea5t th\u00f4ng tin \u0111\u1ed1i s\u00e1nh gi\u1eefa 2 ng\u00f4n ng\u1eef.<\/span><\/p>\n

C\u00e1c n<\/span>g\u1eef li\u1ec7u <\/span>thu th\u1eadp \u0111\u01b0\u1ee3c s\u1ebd<\/span> ch\u01b0a <\/span>c\u00f3<\/span> ch\u00fa th\u00edch th\u00f4ng tin ng\u00f4n ng\u1eef (ng\u1eef li\u1ec7u th\u00f4). <\/span>Trong <\/span>nghi\u00ean c\u1ee9u<\/span> n\u00e0y, ch\u00fang t\u00f4i x\u00e2y d\u1ef1ng ng\u1eef li\u1ec7u <\/span>c\u00f3 ch\u00fa th\u00edch<\/span> nh\u1eb1m sau n\u00e0y c\u00f3 th\u1ec3<\/span> khai th\u00e1c \u0111\u01b0\u1ee3c nhi\u1ec1u tri th\u1ee9c ng\u00f4n ng\u1eef<\/span> h\u01a1n.<\/span> T<\/span>h\u00f4ng tin ng\u00f4n ng\u1eef \u0111\u01b0\u1ee3c ch\u00fa th\u00edch <\/span>(hay c\u00f2n g\u1ecdi l\u00e0 nh\u00e3n ng\u00f4n ng\u1eef) c\u00f3 th\u1ec3 <\/span>l\u00e0 th\u00f4ng tin v\u1ec1 <\/span>b\u00ecnh<\/span> di\u1ec7n <\/span>h\u00ecnh th\u00e1i<\/span>,<\/span> ng\u1eef ph\u00e1p<\/span> v<\/span>\u00e0 <\/span>ng\u1eef ngh\u0129a<\/span> c\u1ee7a <\/span>c\u00e1c \u0111\u01a1n v\u1ecb ng\u00f4n ng\u1eef nh\u01b0 <\/span>t\u1eeb<\/span>, ng\u1eef, c\u00e2u<\/span>. <\/span>Trong <\/span>nghi\u00ean c\u1ee9u<\/span> n\u00e0y, b\u01b0\u1edbc \u0111\u1ea7u, ch\u00fang t\u00f4i ch\u1ec9 m\u1edbi g\u00e1n <\/span>nh\u00e3n <\/span>h\u00ecnh th\u00e1i<\/span>, <\/span>nh\u00e3n <\/span>ng\u1eef ph\u00e1p<\/span> v\u00e0 <\/span>nh\u00e3n <\/span>ng\u1eef ngh\u0129a<\/span> cho <\/span>\u0111\u01a1n v\u1ecb<\/span> t\u1eeb<\/span>. <\/span>Nh\u00e3n h\u00ecnh th\u00e1i t\u1eeb<\/span> \u1edf<\/span> \u0111\u00e2y <\/span>ch\u00ednh l\u00e0<\/span> nh\u00e3n<\/span> ranh gi\u1edbi t\u1eeb<\/span>.<\/span> Nh\u00e3n ng\u1eef ph\u00e1p t\u1eeb<\/span> \u1edf \u0111\u00e2y bao g\u1ed3m c\u00e1c nh\u00e3n ph\u00e2n lo\u1ea1i c\u0103n c\u1ee9 theo m\u1eb7t ng\u1eef ph\u00e1p c\u1ee7a t\u1eeb (hay c\u00f2n g\u1ecdi l\u00e0 <\/span>t\u1eeb ph\u00e1p<\/span>), c\u1ee5 th\u1ec3 bao g\u1ed3m hai ph\u1ea1m tr\u00f9 ng\u1eef ph\u00e1p c\u1ee7a t\u1eeb: <\/span>ph\u1ea1m tr\u00f9 ph\u00e2n lo\u1ea1i t\u1eeb<\/span> v\u00e0 <\/span>ph\u1ea1m tr\u00f9 ng\u1eef ph\u00e1p bi\u1ebfn \u0111\u1ed5i t\u1eeb<\/span>.<\/span><\/p>\n

Kh\u00e1c v\u1edbi nh\u00e3n h\u00ecnh th\u00e1i v\u00e0 ng\u1eef ph\u00e1p (\u0111\u01a1n gi\u1ea3n v\u00e0 d\u1ec5 th\u1ed1ng nh\u1ea5t v\u1edbi nhau), nh\u00e3n ng\u1eef ngh\u0129a hi\u1ec7n c\u00f2n nhi\u1ec1u tranh c\u00e3i v\u1ec1 c\u00e1ch ph\u00e2n lo\u1ea1i. <\/span>Qua kh\u1ea3o s\u00e1t \u00fd ngh\u0129a t\u1eeb v\u1ef1ng c\u1ee7a m\u1ed7i <\/span>th\u1ef1c t\u1eeb<\/span>, ta th\u1ea5y n\u00f3i chung m\u1ed7i t\u1eeb c\u00f3 th\u1ec3 mang nhi\u1ec1u ngh\u0129a <\/span>t\u1eeb v\u1ef1ng <\/span>kh\u00e1c nhau, nh\u01b0ng trong m\u1ed9t ng\u1eef c\u1ea3nh c\u1ee5 th\u1ec3, ch\u00fang s\u1ebd mang m\u1ed9t ngh\u0129a nh\u1ea5t \u0111\u1ecbnh n\u00e0o \u0111\u00f3. Ch\u1eb3ng h\u1ea1n, danh t\u1eeb \u201cbank\u201d trong ti\u1ebfng Anh c\u00f3 th\u1ec3 l\u00e0 \u201cng\u00e2n h\u00e0ng\u201d, <\/span>\u201ckho l\u01b0u tr\u1eef\u201d,<\/span> \u201cb\u1edd s\u00f4ng\u201d<\/span>,<\/span> \u201cd\u00e3y\u201d<\/span>, …<\/span>; danh t\u1eeb \u201c\u0111\u01b0\u1eddng\u201d trong ti\u1ebfng Vi\u1ec7t c\u00f3 th\u1ec3 c\u00f3 ngh\u0129a l\u00e0 \u201c\u0111\u01b0\u1eddng \u0103n\u201d (sugar) hay \u201c\u0111\u01b0\u1eddng \u0111i\u201d (<\/span>street<\/span>), \u2026 \u0110\u1ec3 d\u1ec5 ph\u00e2n bi\u1ec7t c\u00e1c ngh\u0129a t\u1eeb v\u1ef1ng kh\u00e1c nhau, c\u00e1c nh\u00e0 ng\u1eef ngh\u0129a h\u1ecdc, t\u1eeb v\u1ef1ng h\u1ecdc v\u00e0 t\u00e2m l\u00fd h\u1ecdc \u2013 ng\u00f4n ng\u1eef \u0111\u00e3 ph\u00e2n chia to\u00e0n b\u1ed9 c\u00e1c \u00fd ngh\u0129a t\u1eeb v\u1ef1ng c\u00f3 th\u1ec3 c\u00f3 th\u00e0nh h\u1ec7 th\u1ed1ng c\u00e1c \u00fd ni\u1ec7m (c\u00e2y \u00fd ni\u1ec7m) v\u00e0 m\u1ed7i \u00fd ni\u1ec7m nh\u01b0 v\u1eady \u0111\u01b0\u1ee3c coi nh\u01b0 l\u00e0 m\u1ed9t <\/span>l\u1edbp ng\u1eef ngh\u0129a hay <\/span>nh\u00e3n ng\u1eef ngh\u0129a<\/span> c\u1ee7a t\u1eeb<\/span>.<\/span> \u0110\u1ebfn nay, \u0111\u00e3 c\u00f3 m\u1ed9t s\u1ed1 c\u00e1ch ph\u00e2n l\u1edbp <\/span>ng\u1eef ngh\u0129a<\/span> cho ti\u1ebfng Anh\u00a0<\/span>[2]<\/span>, nh\u01b0: LLOCE<\/span> (Longman Lexicon of Contemporary <\/span>E<\/span>nglish) g\u1ed3m 2500 l\u1edbp<\/span>, WordNet<\/span> (kho\u1ea3ng 110.000 l\u1edbp)<\/span>, CoreLex<\/span> (126 l\u1edbp)<\/span>, … <\/span>V\u00ed d\u1ee5, theo h\u1ec7 th\u1ed1ng CoreLex, <\/span>c\u00e1c ngh\u0129a t\u01b0\u01a1ng \u1ee9ng c\u1ee7a <\/span>danh t\u1eeb \u201cbank\u201d<\/span> s\u1ebd l\u00e0: \u201cng\u00e2n h\u00e0ng\u201d thu\u1ed9c v\u1ec1 \u00fd ni\u1ec7m \u201cc\u00f4ng tr\u00ecnh nh\u00e2n t\u1ea1o\u201d (nh\u00e3n <\/span>ART<\/span>); \u201cb\u1edd s\u00f4ng\u201d s\u1ebd thu\u1ed9c v\u1ec1 \u00fd ni\u1ec7m \u201cc\u00f4ng tr\u00ecnh thi\u00ean t\u1ea1o\u201d (nh\u00e3n NAT); \u201cd\u00e3y\u201d s\u1ebd thu\u1ed9c v\u1ec1 \u00fd ni\u1ec7m \u201cs\u1ef1 s\u1eafp x\u1ebfp t\u1ed5 ch\u1ee9c\u201d (nh\u00e3n GRP). T\u01b0\u01a1ng t\u1ef1 cho danh t\u1eeb \u201c\u0111\u01b0\u1eddng\u201d trong ti\u1ebfng Vi\u1ec7t, ngh\u0129a \u201c\u0111\u01b0\u1eddng \u0103n\u201d s\u1ebd \u0111\u01b0\u1ee3c x\u1ebfp v\u00e0o \u00fd ni\u1ec7m \u201c<\/span>h\u00f3a<\/span> ch\u1ea5t\u201d (nh\u00e3n CHM); c\u00f2n ngh\u0129a \u201c\u0111\u01b0\u1eddng \u0111i\u201d s\u1ebd \u0111\u01b0\u1ee3c x\u1ebfp v\u00e0o \u00fd ni\u1ec7<\/span>m \u201ckh\u00f4ng gian\u201d (nh\u00e3n SPA<\/span>); \u2026<\/span><\/p>\n

N<\/span>g\u1eef li\u1ec7u song ng\u1eef <\/span>th\u00f4 (ch\u01b0a qua x\u1eed l\u00fd) <\/span>c\u00f3 th\u1ec3 \u0111\u01b0\u1ee3c x\u00e2y d\u1ef1ng b\u1eb1ng <\/span>3 <\/span>c\u00e1ch<\/span> ch\u00ednh<\/span>:<\/span> (1) <\/span>T<\/span>hu th\u1eadp t\u1ef1 \u0111\u1ed9ng t\u1eeb c\u00e1c <\/span>website song ng\u1eef<\/span>\u00a0 <\/span>(2) <\/span>Thu th\u1eadp t<\/span>\u1eeb <\/span>c\u00e1c \u1ea5n ph\u1ea9m song ng\u1eef <\/span>(d\u1ea1ng \u0111i\u1ec7n t\u1eed)\u00a0 (3) <\/span>D<\/span>\u1ecbch th\u1ee7 c\u00f4ng<\/span>,<\/span> d\u1ecbch song song 1-1 theo h\u01b0\u1edbng d\u1eabn (guideline) <\/span>t\u1eeb c\u00e1c v\u0103n b\u1ea3n ngu\u1ed3n c\u00f3 ch\u1ea5t l\u01b0\u1ee3ng<\/span> v\u00e0<\/span> \u0111\u00fang l\u0129nh v\u1ef1c<\/span>, ni\u00ean \u0111\u1ea1i<\/span>. C\u00e1ch <\/span>(1)<\/span> nhanh, chi ph<\/span>\u00ed th\u1ea5p<\/span>,<\/span> s\u1ed1 l\u01b0\u1ee3ng l\u1edbn<\/span>, ch\u1ea5t l\u01b0\u1ee3ng th\u1ea5p (v\u00ec th\u01b0\u1eddng<\/span> kh\u00f4ng \u0111\u01b0\u1ee3c d\u1ecbch song song 1-1<\/span>),<\/span> kh\u00f4ng <\/span>ph\u00f9 h\u1ee3p l\u0129nh v\u1ef1c v\u00e0 ch\u01b0a gi\u00f3ng h\u00e0ng \u1edf m\u1ee9c n\u00e0o<\/span>. <\/span>C<\/span>\u00e1ch <\/span>(2) <\/span>chi ph\u00ed th\u1ea5p, <\/span>s\u1ed1 l\u01b0\u1ee3ng \u00edt<\/span>, <\/span>ch\u1ea5t l\u01b0\u1ee3ng <\/span>trung b\u00ecnh v<\/span>\u00e0 ch\u01b0a gi\u00f3ng h\u00e0ng c\u00e2u<\/span>. C\u00e1ch (3) <\/span>ch\u1ea5t l\u01b0\u1ee3ng cao, l\u0129nh v\u1ef1c <\/span>ph\u00f9 h\u1ee3p, \u0111\u00e3 gi\u00f3ng h\u00e0ng c\u00e2u<\/span>, nh\u01b0ng chi ph\u00ed cao<\/span>.<\/span> Trong <\/span>nghi\u00ean c\u1ee9u<\/span> n\u00e0y, n<\/span>g\u1eef li\u1ec7u song ng\u1eef Anh-Vi\u1ec7t <\/span>\u0111\u01b0\u1ee3c tr\u00edch m\u1ed9t ph\u1ea7n t\u1eeb kho ng\u1eef li\u1ec7u song ng\u1eef EVC<\/span> (ph\u1ea7n m\u1ec1m [1])<\/span> do <\/span>c\u00e1c nh\u00e0 nghi\u00ean c\u1ee9u c\u1ee7a <\/span>Trung t\u00e2m <\/span>Ng\u00f4n ng\u1eef h\u1ecdc T\u00ednh to\u00e1n c\u1ee7a Tr\u01b0\u1eddng \u0110H Khoa h\u1ecdc T\u1ef1 nhi\u00ean<\/span> –<\/span> Tp.HCM x\u00e2y d\u1ef1ng<\/span> (b\u1eb1ng c\u00e1ch 2 v\u00e0 3) <\/span>v\u00e0 <\/span>\u0111\u00e3 <\/span>c\u00f4ng b\u1ed1 \u1edf c\u00f4ng tr\u00ecnh [1]<\/span>.<\/span><\/p>\n

3.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>X\u1eec L\u00dd NG\u1eee LI\u1ec6U SONG NG\u1eee<\/span> ANH-VI\u1ec6T<\/span><\/h1>\n

T\u00f9y theo c\u00e1ch th\u1ee9c x\u00e2y d\u1ef1ng n<\/span>g\u1eef li\u1ec7u song ng\u1eef<\/span> m\u00e0 ta c\u00f3 c\u00e1c c\u00f4ng \u0111o\u1ea1n x\u1eed l\u00fd kh\u00e1c nhau. N\u1ebfu ng\u1eef li\u1ec7u<\/span> \u0111\u01b0\u1ee3c <\/span>thu th\u1eadp <\/span>t\u1eeb website song ng\u1eef, ta ph\u1ea3i gi\u00f3ng h\u00e0ng v\u0103n b\u1ea3n<\/span>, r\u1ed3i sau \u0111\u00f3 g<\/span>i\u00f3ng h\u00e0ng \u1edf m\u1ee9c c\u00e2u (c\u00f3 th\u1ec3 <\/span>s\u1eed d\u1ee5ng c\u00f4ng c\u1ee5 <\/span>ph\u1ea7n m\u1ec1m <\/span>[<\/span>2<\/span>]<\/span>)<\/span>. N\u1ebfu ng\u1eef li\u1ec7u thu th\u1eadp t\u1eeb \u1ea5n ph\u1ea9m song ng\u1eef th\u00ec ch\u1ec9 c\u1ea7n gi\u00f3ng h\u00e0ng c\u00e2u. Ta c\u00f3 <\/span>k\u1ebft qu\u1ea3 <\/span>nh\u01b0 <\/span>minh h\u1ecda <\/span>b\u00ean d\u01b0\u1edbi<\/span> (L\u01b0u \u00fd:<\/span> c\u1ea7n<\/span> ph\u1ea3i chu\u1ea9n h\u00f3a ng\u1eef li\u1ec7u th\u00e0nh d\u1ea1ng text-only v\u00e0 m\u00e3 utf-8<\/span> tr\u01b0\u1edbc khi s\u1eed d\u1ee5ng c\u00e1c c\u00f4ng c\u1ee5 tr\u00ean<\/span>)<\/span>:<\/span><\/p>\n

* Helicopters can rise straight up into the air and can go straight down.<\/span><\/p>\n

+ M\u00e1y bay tr\u1ef1c th\u0103ng c\u00f3 th\u1ec3 l\u00ean th\u1eb3ng tr\u00ean kh\u00f4ng v\u00e0 \u0111\u00e1p th\u1eb3ng xu\u1ed1ng \u0111\u1ea5t.<\/span><\/p>\n

\u00a0<\/span><\/p>\n

* They can stand still in the air.<\/span><\/p>\n

+ Ch\u00fang c\u00f3 th\u1ec3 \u0111\u1ee9ng y\u00ean tr\u00ean kh\u00f4ng.<\/span><\/p>\n

\u00a0<\/span><\/p>\n

* Helicopt<\/span>ers do not have wings.<\/span><\/p>\n

+ M\u00e1y bay tr\u1ef1c th\u0103ng kh\u00f4ng c\u00f3 c\u00e1nh.<\/span><\/p>\n


\n\u0110\u1ec3 gi\u00f3ng h\u00e0ng\u00a0<\/span>t\u1ef1 \u0111\u1ed9ng <\/span>\u1edf m\u1ee9c t\u1eeb<\/span> (word alignment)<\/span>, t<\/span>a c\u00f3 th\u1ec3<\/span> s\u1eed d\u1ee5ng <\/span>c\u00f4ng c\u1ee5 GIZA++ [<\/span>3<\/span>] v\u00e0 \u0111\u1ed9 ch\u00ednh x\u00e1c c\u1ee7a c\u00f4ng c\u1ee5 n\u00e0y t\u00f9y thu\u1ed9c v\u00e0o kh\u1ed1i l\u01b0\u1ee3ng v\u00e0 ch\u1ea5t l\u01b0\u1ee3ng ng\u1eef li\u1ec7u (c\u00e0ng nhi\u1ec1u c\u00e2u song song <\/span>1-1 <\/span>th\u00ec c\u00e1c m\u1ed1i n\u1ed1i c\u00e0ng ch\u00ednh x\u00e1c h\u01a1n).<\/span> Do c\u00f3 s\u1ef1 kh\u00e1c bi\u1ec7t v\u1ec1 lo\u1ea1i h\u00ecnh ng\u00f4n ng\u1eef gi\u1eefa ti\u1ebfng <\/span>Anh (bi\u1ebfn h\u00ecnh) v\u00e0 ti\u1ebfng Vi\u1ec7t (\u0111\u01a1n l\u1eadp), n\u00ean ranh gi\u1edbi t\u1eeb gi\u1eefa 2 ng\u00f4n ng\u1eef l\u00e0 kh\u00e1c nhau. V\u00ec v\u1eady, \u0111\u1ec3 c\u00f3 k\u1ebft<\/span> qu\u1ea3 gi\u00f3ng h\u00e0ng t\u1eeb t\u1ed1t, ch\u00fang t<\/span>a n\u00ean<\/span> ph\u00e2n \u0111o\u1ea1n t\u1eeb (word segmentation) ti\u1ebfng Vi\u1ec7t<\/span> (c\u00f3 th\u1ec3 s\u1eed d\u1ee5ng<\/span> c\u00f4ng c\u1ee5 [<\/span>2<\/span>]<\/span>)<\/span> tr\u01b0\u1edbc <\/span>k<\/span>hi th\u1ef1c hi\u1ec7n vi\u1ec7c gi\u00f3ng h\u00e0ng t\u1eeb (xin xem <\/span>k\u1ebft qu\u1ea3 nh\u01b0 <\/span>h\u00ecnh 2).<\/span><\/p>\n

\"\"\u00a0<\/span>H\u00ecnh 2. C\u1eb7p c\u00e2u \u0111\u00e3 ph\u00e2n \u0111o\u1ea1n t\u1eeb, gi\u00f3ng h\u00e0ng t\u1eeb v\u00e0 g\u00e1n nh\u00e3n t\u1eeb lo\u1ea1i.<\/span><\/p>\n

Ng\u1eef li\u1ec7u song ng\u1eef sau khi \u0111\u01b0\u1ee3c ph\u00e2n \u0111o\u1ea1n t\u1eeb v\u00e0 gi\u00f3ng h\u00e0ng t\u1eeb, c\u00f3 th\u1ec3 \u0111\u01b0\u1ee3c g\u00e1n th\u00eam c\u00e1c th\u00f4ng tin ng\u00f4n ng\u1eef kh\u00e1c, nh\u01b0: t\u1eeb lo\u1ea1i (nh\u00e3n t\u1eeb ph\u00e1p) v\u00e0 ng\u1eef ngh\u0129a. Trong <\/span>nghi\u00ean c\u1ee9u<\/span> n\u00e0y, ch\u00fang t\u00f4i \u0111\u00e3 s\u1eed d\u1ee5ng c\u00f4ng c\u1ee5 [<\/span>2<\/span>] \u0111\u1ec3 g<\/span>\u00e1n nh\u00e3n t\u1eeb lo\u1ea1i cho ti\u1ebfng Vi\u1ec7t <\/span>v\u00e0 c\u00f4ng c\u1ee5 Stanford<\/span> POS<\/span> tagger\u00a0<\/span>[<\/span>4<\/span>] \u0111\u1ec3 g\u00e1n nh\u00e3n t\u1eeb lo\u1ea1i cho ti\u1ebfng Anh. <\/span>Ch\u00fang t\u00f4i <\/span>c\u00f2n<\/span> s\u1eed d\u1ee5ng c\u00f4ng c\u1ee5 <\/span>[2]<\/span> \u0111\u1ec3 g\u00e1n nh\u00e3n ng\u1eef ngh\u0129a cho <\/span>c\u00e1c th\u1ef1c t\u1eeb trong <\/span>song ng\u1eef Anh<\/span> –<\/span> Vi\u1ec7t<\/span> d\u1ef1a tr\u00ean s\u1ef1 r\u00e0ng bu\u1ed9c v\u1ec1 ng\u1eef ngh\u0129a gi\u1eefa c\u00e1c c\u1eb7p t\u1eeb\/c\u1ee5m t\u1eeb t\u01b0\u01a1ng \u1ee9ng (n\u1ebfu l\u00e0 d\u1ecbch 1-1 th\u00ec c\u1ea3 hai ph\u1ea3i c\u00f9ng thu\u1ed9c m\u1ed9t l\u1edbp ng\u1eef ngh\u0129a)<\/span> nh\u01b0 sau:<\/span><\/p>\n

Theo b\u1ed9 nh\u00e3n LLOCE, khi<\/span> ta <\/span>x\u00e9t t\u1eeb \u201cplane\u201d <\/span>v\u00e0 \u201cfly\u201d t\u01b0\u01a1ng \u1ee9ng v\u1edbi t\u1eeb \u201cm\u00e1y bay\u201d v\u00e0 \u201cbay\u201d trong c\u1eb7p c\u00e2u n\u00f3i tr\u00ean, ta th\u1ea5y: \u201cplane\u201d l\u00e0 danh t\u1eeb, s\u1ebd thu\u1ed9c c\u00e1c l\u1edbp J41 (kh\u00f4ng gian ph\u1eb3ng) v\u00e0 M180 (ph\u01b0\u01a1ng ti\u1ec7n h\u00e0ng kh\u00f4ng), c\u00f2n n\u1ebfu \u201cplane\u201d l\u00e0 \u0111\u1ed9ng t\u1eeb s\u1ebd thu\u1ed9c c\u00e1c l\u1edbp kh\u00e1c. T\u01b0\u01a1ng t\u1ef1 cho t\u1eeb \u201cfly\u201d, trong tr\u01b0\u1eddng h\u1ee3p n\u00e0y l\u00e0 \u0111\u1ed9ng t\u1eeb, s\u1ebd thu\u1ed9c l\u1edbp M19 (h\u00e0nh \u0111\u1ed9ng bay). T\u01b0\u01a1ng \u1ee9ng trong ti\u1ebfng Vi\u1ec7t, danh t\u1eeb \u201cm\u00e1y bay\u201d ch\u1ec9 thu\u1ed9c l\u1edbp M180, c\u00f2n \u0111\u1ed9ng t\u1eeb \u201cbay\u201d thu\u1ed9c l\u1edbp M19 v\u00e0 L47 (bay h\u01a1i\/bay m\u00e0u)<\/span>. Do nguy\u00ean l\u00fd chia s\u1ebb c\u00f9ng l\u1edbp ng\u1eef ngh\u0129a, ta x\u00e1c \u0111\u1ecbnh \u0111\u01b0\u1ee3c \u201cplane\u201d v\u00e0 \u201cm\u00e1y bay\u201d c\u00f3 nh\u00e3n ng\u1eef ngh\u0129a l\u00e0 M180 v\u00e0 \u201cfly\u201d v\u00e0 \u201cbay\u201d c\u00f3 c\u00f9ng nh\u00e3n ng\u1eef ngh\u0129a l\u00e0 M19. <\/span><\/p>\n

Do s\u1ef1 kh\u00e1c bi\u1ec7t v\u1ec1 lo\u1ea1i h\u00ecnh ng\u00f4n ng\u1eef, lo\u1ea1i h\u00ecnh v\u0103n h\u00f3a, n\u00ean s\u1ef1 nh\u1eadp nh\u1eb1ng gi\u1eefa 2 ng\u00f4n ng\u1eef <\/span>kh\u00e1c nhau <\/span>th\u01b0\u1eddng l\u00e0 kh\u00e1c nhau. N\u1ebfu khi giao nhau (ph\u00e9p AND), gi\u1eefa 2 t\u1eadp l\u1edbp c\u1ee7a 2 t\u1eeb m\u00e0 k\u1ebft qu\u1ea3 l\u1edbn h\u01a1n 1, l\u00fac \u0111\u00f3 ch\u01b0\u01a1ng tr\u00ecnh d\u00f9ng th\u00eam m\u1ed9t s\u1ed1 th\u00f4ng tin kh\u00e1c \u0111\u1ec3 ch\u1ecdn nh\u00e3n h\u1ee3p l\u00fd nh\u1ea5t. T\u1ea5t nhi\u00ean c\u00e1c k\u1ebft qu\u1ea3 x\u1eed l\u00fd t\u1ef1 \u0111\u1ed9ng tr\u00ean kh\u00f4ng th\u1ec3 ch\u00ednh x\u00e1c ho\u00e0n to\u00e0n<\/span> v\u00e0 c\u0169ng c\u1ea7n s\u1ef1 hi\u1ec7u \u0111\u00ednh th\u1ee7 c\u00f4ng<\/span>. <\/span>V\u1ec1 ti\u00eau ch\u00ed ph\u00e2n \u0111o\u1ea1n t\u1eeb (hay c\u00f2n g\u1ecdi l\u00e0 nh\u1eadn di\u1ec7n ranh gi\u1edbi t\u1eeb) ti\u1ebfng Vi\u1ec7t v\u00e0 b\u1ed9 nh\u00e3n t\u1eeb lo\u1ea1i v\u00e0 b\u1ed9 nh\u00e3n ng\u1eef ngh\u0129a, ch\u00fang t\u00f4i <\/span>k\u1ebf th\u1eeba t\u1eeb<\/span> c\u00f4ng tr\u00ecnh s\u1ed1 [<\/span>2<\/span>].<\/span><\/p>\n

4.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>KHAI TH\u00c1C NG\u1eee LI\u1ec6U SONG NG\u1eee<\/span> ANH-VI\u1ec6T<\/span><\/h1>\n

T\u1eeb kho EVC, ch\u00fang ta c\u00f3 th\u1ec3 khai th\u00e1c \u0111\u1ec3 ph\u1ee5c v\u1ee5 cho r\u1ea5t nhi\u1ec1u b\u00e0i to\u00e1n \u1edf c\u00e1c<\/span> l\u0129nh v\u1ef1c kh\u00e1c nhau<\/span>, nh\u01b0<\/span>:<\/span> th\u1ed1ng k\u00ea <\/span>ng\u00f4n ng\u1eef<\/span>, <\/span>so s\u00e1nh \u0111\u1ed1i chi\u1ebfu<\/span> ng\u00f4n ng\u1eef <\/span>\u0111\u1ec3 ph\u1ee5c v\u1ee5 cho vi\u1ec7c nghi\u00ean c\u1ee9u, gi\u1ea3ng d\u1ea1y ng\u00f4n ng\u1eef<\/span>;<\/span> ph\u00e1t hi\u1ec7n quy lu\u1eadt ng\u00f4n ng\u1eef<\/span>,<\/span>\u2026 K<\/span>ho ng\u1eef li\u1ec7u c\u00e0ng <\/span>l\u1edbn v\u00e0 <\/span>\u0111\u01b0\u1ee3c g\u00e1n nhi\u1ec1u th\u00f4ng tin ng\u00f4n ng\u1eef th\u00ec hi\u1ec7u qu\u1ea3 c\u1ee7a vi\u1ec7c khai th\u00e1c <\/span>c\u00e0ng l\u1edbn<\/span>. D\u01b0\u1edbi \u0111\u00e2y l\u00e0 m\u1ed9t s\u1ed1 v\u00ed d\u1ee5 khai th\u00e1c EVC nh\u1eb1m ph\u1ee5c v\u1ee5 gi\u1ea3ng d\u1ea1y ng\u00f4n ng\u1eef:<\/span><\/p>\n

4.1.<\/span>\u00a0 \u00a0 \u00a0 \u00a0 \u00a0<\/span>Th\u1ed1ng k\u00ea <\/span>ng\u00f4n ng\u1eef<\/span><\/h2>\n

4.1.1.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>Th\u1ed1ng k\u00ea theo h\u00ecnh th\u00e1i t\u1eeb<\/span>:<\/span><\/h3>\n

Do \u0111\u1eb7c th\u00f9 c\u1ee7a ti\u1ebfng Vi\u1ec7t, n\u00ean khi ch\u00fang ta s\u1eed d\u1ee5ng c\u00e1c c\u00f4ng c<\/span>\u1ee5<\/span> t\u00ecm ki\u1ebfm, th\u1ed1ng k\u00ea ng\u00f4n ng\u1eef c\u1ee7a <\/span>ti\u1ebfng Anh<\/span>, ch\u00fang ta s\u1ebd kh\u00f4ng th\u1ec3 x\u00e1c \u0111\u1ecbnh \u0111\u00fang \u0111\u01b0\u1ee3c h\u00ecnh th\u00e1i c\u1ee7a chu\u1ed7i \u0111ang t\u00ecm<\/span> (v\u00ec ch\u00fang xem ti\u1ebfng l\u00e0 t\u1eeb)<\/span>. <\/span>C\u00f2n trong ng\u1eef li\u1ec7u EVC, do<\/span> c\u00f3 g\u00e1n nh\u00e3n h\u00ecnh th\u00e1i t\u1eeb, <\/span>n\u00ean<\/span> vi\u1ec7c t\u00ecm ki\u1ebfm ti\u1ebfng Vi\u1ec7t s\u1ebd hi\u1ec7u qu\u1ea3 h\u01a1n. V\u00ed d\u1ee5 ta mu\u1ed1n t\u00ecm <\/span>t\u1eeb \u201ctin\u201d: m\u00e1y s\u1ebd t\u00ecm ra t\u1eeb \u201ctin\u201d n\u1eb1m \u0111\u1ed9c l\u1eadp (nh\u01b0: \u201c<\/span>tin <\/span>\u0111i\u1ec1u \u0111\u00f3\u2026\u201d, \u201c<\/span>tin<\/span> m\u1edbi nh\u1eadn\u201d), ho\u1eb7c t\u1eeb \u201ctin\u201d trong ng\u1eef: \u201cnh\u1eafn <\/span>tin<\/span>\u201d, \u201c<\/span>tin<\/span> s\u1ed1t d\u1ebbo\u201d, \u2026; ch\u1ee9 <\/span>m\u00e1y <\/span>kh\u00f4ng b\u1ecb nh\u1ea7m l\u1eabn v\u1edbi <\/span>h\u00ecnh v\u1ecb <\/span>\u201ctin\u201d trong c\u00e1c <\/span>t\u1eeb<\/span> \u201ctin <\/span>m\u1eebng<\/span>\u201d<\/span>,<\/span> \u201ctin t\u1ee9c\u201d, \u201cth\u00f4ng tin\u201d hay <\/span>\u00e1-h\u00ecnh v\u1ecb<\/span> \u201ctin\u201d trong \u201cc\u0103n-tin\u201d, \u2026<\/span>T\u01b0\u01a1ng t\u1ef1, khi t\u00ecm t\u1eeb \u201cquan t\u00e0i\u201d, m\u00e1y<\/span> s\u1ebd kh\u00f4ng nh\u1ea7m <\/span>v\u1edbi <\/span>c<\/span>\u1ee5m<\/span> \u201cq<\/span>uan t\u00e0i\u201d trong c\u00e2u \u201cm\u1ed9t \u00f4ng <\/span>quan t\u00e0i<\/span> gi\u1ecfi\u201d.<\/span> T<\/span>\u01b0\u01a1ng t\u1ef1, ch\u00fang ta c\u00f3 th\u1ec3 t\u00ecm <\/span>t\u1eeb<\/span> ti\u1ebfng Anh <\/span>ch\u00ednh x\u00e1c ho\u1eb7c t\u1ea5t c\u1ea3<\/span> c\u00e1c bi\u1ebfn c\u00e1ch <\/span>(inflection) <\/span>c\u1ee7a<\/span> n\u00f3, nh\u01b0: t\u1eeb \u201cdisplay\u201d, \u201cdisplays\u201d, \u201cdisplayed\u201d hay \u201cdisplaying\u201d.<\/span><\/p>\n

4.1.2.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>Th\u1ed1ng k\u00ea <\/span>theo <\/span>ng\u1eef <\/span>ph\u00e1p<\/span> t\u1eeb<\/span>:<\/span><\/h3>\n

Ch\u00fang ta c\u00f3 th\u1ec3 t\u00ecm ki\u1ebfm t\u1eeb theo t\u1eeb lo\u1ea1i c\u1ee7a n\u00f3, v\u00ed d\u1ee5: t<\/span>\u00ecm \u0111\u1ed9ng t\u1eeb \u201ctin\u201d: m\u00e1y s\u1ebd t\u00ecm ra \u0111\u00fang \u0111\u1ed9ng t\u1eeb \u201ctin\u201d n\u1eb1m \u0111\u1ed9c l\u1eadp trong <\/span>c\u00e1c tr\u01b0\u1eddng h\u1ee3p nh\u01b0: <\/span>\u201cch\u00fang ta <\/span>tin <\/span>r\u1eb1ng \u2026\u201d; ho\u1eb7c <\/span>t\u00ecm <\/span>danh t\u1eeb \u201ctin\u201d trong <\/span>c\u00e1c <\/span>ng\u1eef: \u201cnh\u1eafn <\/span>tin<\/span>\u201d, \u201c<\/span>tin<\/span> s\u1ed1t d\u1ebbo\u201d, \u2026; <\/span><\/p>\n

T\u01b0\u01a1ng t\u1ef1 cho vi\u1ec7c t\u00ecm<\/span> \u0111\u1ed9ng t\u1eeb \u201cdisplay\u201d: m\u00e1y s\u1ebd t\u00ecm ra t\u1eeb \u0111\u00f3 trong \u201c<\/span>display<\/span> information\u201d (c\u0169ng nh\u01b0 c\u00e1c bi\u1ebfn c\u00e1ch \u0111\u1ed9ng t\u1eeb c\u1ee7a n\u00f3, nh\u01b0: \u201c<\/span>displayed<\/span> information\u201d, \u201c<\/span>displays<\/span> information\u201d, \u201c<\/span>displaying<\/span> information\u201d); ho\u1eb7c <\/span>ch\u1ec9 t\u00ecm <\/span>danh t\u1eeb \u201cdisplay\u201d trong \u201ca new <\/span>display<\/span>\u201d (c\u0169ng nh\u01b0 c\u00e1c bi\u1ebfn c\u00e1ch danh t\u1eeb c\u1ee7a n\u00f3, nh\u01b0: \u201cmany new displays\u201d ).<\/span><\/p>\n

V\u1edbi th\u00f4ng tin v\u1ec1 <\/span>ti\u1ec3u<\/span> t\u1eeb lo\u1ea1i <\/span>v\u00e0<\/span> ng\u1eef ph\u00e1p bi\u1ebfn \u0111\u1ed5i t\u1eeb<\/span>, kho ng\u1eef li\u1ec7u EVC c\u00f3 th\u1ec3 \u0111\u00e1p \u1ee9ng \u0111\u01b0\u1ee3c c\u00e1c y\u00eau c\u1ea7u khai th\u00e1c chi ti\u1ebft h\u01a1n, nh\u01b0:<\/span> t<\/span>\u00ecm <\/span>t\u1eeb<\/span> ti\u1ebfng Vi\u1ec7t theo ti\u1ec3u t\u1eeb lo\u1ea1i<\/span> (<\/span>\u0111\u1ed9ng t\u1eeb n\u1ed9i \u0111\u1ed9ng; ngo\u1ea1i \u0111\u1ed9ng; danh t\u1eeb \u0111\u01a1n th\u1ec3, danh t\u1eeb t\u1ed5ng th\u1ec3, danh t\u1eeb kh\u1ed1i, \u2026<\/span>); t<\/span>\u00ecm <\/span>t\u1eeb<\/span> ti\u1ebfng Anh theo ti\u1ec3u t\u1eeb lo\u1ea1i v\u00e0 ng\u1eef ph\u00e1p bi\u1ebfn \u0111\u1ed5i t\u1eeb<\/span> (<\/span>nh\u01b0: ch\u1ec9 t\u00ecm nh\u1eefng \u0111\u1ed9ng t\u1eeb<\/span> ng\u00f4<\/span>i 3 s\u1ed1 \u00edt, danh t\u1eeb s\u1ed1 nhi\u1ec1u, \u2026); <\/span>Ngo\u00e0i ra, ch\u00fang ta c\u0169ng c\u00f3 th\u1ec3 th\u1ed1ng k\u00ea xem trong kho ng\u1eef li\u1ec7u c\u1ee7a ch\u00fang ta: m\u1ed7i <\/span>t\u1eeb lo\u1ea1i\/ti\u1ec3u t\u1eeb lo\u1ea1i<\/span> c\u00f3 <\/span>bao nhi\u00eau <\/span>t\u1eeb<\/span>, bao nhi\u00eau <\/span>l\u01b0\u1ee3t t\u1eeb<\/span> cho m\u1ed7i ng\u00f4n ng\u1eef.<\/span><\/p>\n

4.1.3.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>Th\u1ed1ng k\u00ea <\/span>theo ng\u1eef ngh\u0129a t\u1eeb<\/span>:<\/span><\/h3>\n

Ta c\u00f3 th\u1ec3 th\u1ed1ng k\u00ea ri\u00eang <\/span>danh t\u1eeb \u201c\u0111\u01b0\u1eddng\u201d <\/span>(sugar) <\/span>trong ng\u1eef li\u1ec7u ti\u1ebfng Vi\u1ec7t, <\/span>tr\u00e1nh b\u1ecb nh\u1ea7m l\u1eabn v\u1edbi c\u00e1c danh t\u1eeb \u201c\u0111\u01b0\u1eddng\u201d kh\u00e1c (<\/span>nh\u01b0: <\/span>street, line, …) b\u1eb1ng c\u00e1ch t\u00ecm danh t\u1eeb \u201c\u0111\u01b0\u1eddng\u201d v\u1edbi nh\u00e3n ng\u1eef ngh\u0129a l\u00e0 CHM (chemicals). T\u01b0\u01a1ng t\u1ef1, ta<\/span> c\u00f3 th\u1ec3 t\u00ecm t\u1ef1 \u0111\u1ed9ng r\u1ea5t nhanh t\u1ea5t c\u1ea3 c\u00e1c t\u1eeb theo m\u1ed9t ch\u1ee7 \u0111\u1ec1 c\u1ee5 th\u1ec3 n\u00e0o \u0111\u00f3 m\u00e0 c\u00e1c ph\u01b0\u01a1ng ph\u00e1p t\u00ecm ki\u1ebfm truy\u1ec1n th\u1ed1ng tr\u01b0\u1edbc \u0111\u00e2y kh\u00f4ng th\u1ec3 th\u1ef1c hi\u1ec7n \u0111\u01b0\u1ee3c. Ch\u1eb3ng h\u1ea1n ta mu\u1ed1n t\u00ecm <\/span>t\u1ea5t c\u1ea3<\/span> c\u00e1c <\/span>t\u1eeb n\u00f3i v\u1ec1 truy\u1ec1n th\u00f4ng (COM)<\/span> trong ng\u1eef li\u1ec7u<\/span> (nh\u01b0 h\u00ecnh <\/span>3<\/span>)<\/span>. <\/span>T\u01b0\u01a1ng t\u1ef1 cho c\u00e1c y\u00eau c\u1ea7u t\u00ecm ki\u1ebfm danh t\u1eeb v\u1ec1 <\/span>th\u1ee9c \u0103n <\/span>(d\u00f9ng nh\u00e3n<\/span> E1<\/span>),<\/span> b\u1ed9 ph\u1eadn c\u01a1 th\u1ec3 con ng\u01b0\u1eddi (nh\u00e3n C3)<\/span>, \u2026<\/span><\/p>\n

\"\"<\/p>\n

H\u00ecnh <\/span>3<\/span>. <\/span>T\u00ecm c\u00e1c <\/span>t<\/span>\u1eeb v\u1ec1 \u201ctruy\u1ec1n th\u00f4ng\u201d (COMmunication).<\/span><\/p>\n

Ngo\u00e0i nh\u1eefng th\u1ed1ng k\u00ea ch\u00ednh n\u00eau tr\u00ean, ch\u00fang ta c\u00f2n c\u00f3 th\u1ec3 t\u00ecm ki\u1ebfm, th\u1ed1ng k\u00ea theo nhi\u1ec1u th\u00f4ng s\u1ed1 kh\u00e1c v\u00e0 c\u00f3 th\u1ec3 k\u1ebft h\u1ee3p c\u00e1c tham s\u1ed1 \u0111\u00f3 v\u1edbi nhau. V\u00ed d\u1ee5: <\/span>th\u1ed1ng k\u00ea v\u1ec1 chi\u1ec1u d\u00e0i c\u00e2u, t\u1ea7n su\u1ea5t s\u1eed d\u1ee5ng \u0111\u1ed9ng t\u1eeb\/danh t\u1eeb, t\u1ea7n su\u1ea5t s\u1eed d\u1ee5ng m\u1ed9t t\u1eeb c\u1ee5 th\u1ec3 <\/span>hay<\/span> c\u00e1c t\u1eeb c\u00f9ng l\u1edbp ng\u1eef ngh\u0129a<\/span>. K\u1ebft qu\u1ea3 th\u1ed1ng k\u00ea n\u00e0y c\u00f3 th\u1ec3 \u0111\u01b0\u1ee3c d\u00f9ng \u0111\u1ec3: <\/span>x\u00e2y d\u1ef1ng v\u1ed1n t\u1eeb c\u01a1 b\u1ea3n, s<\/span>o s\u00e1nh \u0111\u1ed9 t\u01b0\u01a1ng \u0111\u1ed3ng v\u0103n b\u1ea3n, truy t\u00ecm t\u00e1c gi\u1ea3 khuy\u1ebft danh, kh\u00e1m ph\u00e1 phong c\u00e1ch c\u1ee7a t\u00e1c gi\u1ea3, \u0111o\u00e1n nh\u1eadn tr\u1ecdng t\u00e2m t\u00e1c ph\u1ea9m<\/span>, ki\u1ec3m nghi\u1ec7m gi\u1ea3 thuy\u1ebft trong ng\u00f4n ng\u1eef<\/span>, \u2026<\/span><\/p>\n

M\u1ed9t \u0111i\u1ec1u \u0111\u1eb7c bi\u1ec7t l\u00e0 t\u1ea5t c\u1ea3 c\u00e1c k\u1ebft qu\u1ea3 t\u00ecm ki\u1ebfm trong ti\u1ebfng Vi\u1ec7t s\u1ebd c\u00f3 c\u00e2u ti\u1ebfng Anh hi\u1ec3n th\u1ecb song song b\u00ean d\u01b0\u1edbi v\u00e0 \u0111\u00e1nh d\u1ea5u t\u1eeb\/c\u1ee5m t\u1eeb ti\u1ebfng Anh t\u01b0\u01a1ng \u1ee9ng v\u1edbi t\u1eeb\/c\u1ee5m t\u1eeb ti\u1ebfng Vi\u1ec7t \u0111\u00f3 v\u00e0 ng\u01b0\u1ee3c l\u1ea1i <\/span>(<\/span>v\u00ec EVC c\u1ee7a ch\u00fang ta \u0111\u00e3 \u0111\u01b0\u1ee3c gi\u00f3ng h\u00e0ng c\u00e2u v\u00e0 gi\u00f3ng h\u00e0ng t\u1eeb<\/span>)<\/span>. \u0110\u00e2y l\u00e0 \u0111i\u1ec1u v\u00f4 c\u00f9ng c\u1ea7n thi\u1ebft <\/span>khi<\/span> so s\u00e1nh, \u0111\u1ed1i chi\u1ebfu gi\u1eefa 2 ng\u00f4n ng\u1eef.<\/span><\/p>\n

4.2.<\/span>\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>So s\u00e1nh \u0111\u1ed1i chi\u1ebfu ng\u00f4n ng\u1eef<\/span><\/h2>\n

Kho EVC s\u1ebd gi\u00fap ch\u00fang ta <\/span>so s\u00e1nh \u0111\u1ed1i chi\u1ebfu c\u00e1c \u0111i\u1ec3m t\u01b0\u01a1ng \u0111\u1ed3ng v\u00e0 d\u1ecb bi\u1ec7t<\/span>, nh\u01b0<\/span>:<\/span><\/p>\n

4.2.1.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>So s\u00e1nh Anh \u2013 Vi\u1ec7t v\u1ec1 s\u1ef1 t\u1eeb v\u1ef1ng <\/span>h\u00f3a<\/span>:<\/span><\/h3>\n

Quan s\u00e1t c\u00e1c<\/span> m\u1ed1i n\u1ed1i t\u1eeb t\u01b0\u01a1ng \u1ee9ng trong EVC<\/span> (h\u00ecnh 2)<\/span>, ta <\/span>s\u1ebd <\/span>th\u1ea5y s<\/span>\u1ef1<\/span> kh\u00e1c bi\u1ec7t v\u1ec1 t\u1eeb v\u1ef1ng <\/span>h\u00f3a<\/span> gi\u1eefa 2 ng\u00f4n ng\u1eef<\/span>: c\u00f3 <\/span>kh\u00e1i ni\u1ec7m<\/span> \u0111\u01b0\u1ee3c t\u1eeb v\u1ef1ng <\/span>h\u00f3a<\/span> trong ng\u00f4n ng\u1eef n\u00e0y nh\u01b0ng l\u1ea1i kh\u00f4ng \u0111\u01b0\u1ee3c t\u1eeb v\u1ef1ng <\/span>h\u00f3a<\/span> trong ng\u00f4n ng\u1eef kia v\u00e0 ng\u01b0\u1ee3c l\u1ea1i. V\u00ed d\u1ee5: \u201ccow\u201d \u0111\u1ec3 ch\u1ec9 \u201cb\u00f2 c\u00e1i\u201d \u0111\u01b0\u1ee3c xem l\u00e0 <\/span>t\u1eeb<\/span> trong ti\u1ebfng Anh nh\u01b0ng l\u1ea1i kh\u00f4ng l\u00e0 <\/span>t\u1eeb<\/span> trong ti\u1ebfng Vi\u1ec7t<\/span> (l\u00fac n\u00e0y m\u1ed1i n\u1ed1i gi\u00f3ng h\u00e0ng t\u1eeb kh\u00f4ng c\u00f2n l\u00e0 <\/span>1-1<\/span> m\u00e0 l\u00e0 <\/span>1-n<\/span> hay <\/span>m-1<\/span>).<\/span><\/p>\n

4.2.2.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>So s\u00e1nh Anh \u2013 Vi\u1ec7t v\u1ec1 t\u1eeb <\/span>lo\u1ea1i<\/span>:<\/span><\/h3>\n

Quan s\u00e1t<\/span> nh\u00e3n t\u1eeb lo\u1ea1i <\/span>v\u00e0 m\u1ed1i n\u1ed1i t\u1eeb <\/span>trong EVC<\/span> (h\u00ecnh 2)<\/span>, <\/span>ta s\u1ebd th\u1ea5y <\/span>c\u00f3 khi <\/span>trong ti\u1ebfng Anh d\u00f9ng <\/span>danh <\/span>h\u00f3a<\/span> (nominalization) c\u00f2n ti\u1ebfng Vi\u1ec7t <\/span>t\u01b0\u01a1ng <\/span>\u1ee9<\/span>ng l\u1ea1i<\/span> d\u00f9ng <\/span>\u0111\u1ed9ng <\/span>h\u00f3a<\/span> (verbalization). <\/span>C\u00f3 ngh\u0129a l\u00e0 \u0111\u00f4i khi 2 b\u00ean kh\u00f4ng c\u00f9ng t\u1eeb lo\u1ea1i. <\/span>V\u00ed d\u1ee5: trong ti\u1ebfng Anh, ng\u01b0\u1eddi ta n\u00f3i \u201cthank you for your <\/span>attention<\/span>\u201d<\/span>, ti\u1ebfng Vi\u1ec7t l\u00e0<\/span>: <\/span>\u201c<\/span>c\u00e1m \u01a1n c\u00e1c b\u1ea1n<\/span> \u0111\u00e3 ch\u00fa \u00fd<\/span>\u201d.<\/span><\/p>\n

4.2.3.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>So s\u00e1nh Anh \u2013 Vi\u1ec7t v\u1ec1 t<\/span>r\u1eadt t\u1ef1 t<\/span>\u1eeb:<\/span><\/h3>\n

Do <\/span>EVC<\/span> \u0111\u00e3 \u0111\u01b0\u1ee3c gi\u00f3ng h\u00e0ng t\u1eeb<\/span>, <\/span>n\u00ean khi quan s\u00e1t c\u00e1c m\u1ed1i n\u1ed1i (h\u00ecnh <\/span>2<\/span>), ta s\u1ebd th\u1ea5y <\/span>c\u00f3 s\u1ef1<\/span> thay \u0111\u1ed5i v\u1ec1 tr\u1eadt t\u1ef1 t\u1eeb<\/span> gi\u1eefa 2 ng\u00f4n ng\u1eef<\/span> (c\u00e1c m\u1ed1i n\u1ed1i b\u1ecb ch\u00e9o nhau).<\/span><\/p>\n

4.2.4.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>So s\u00e1nh Anh \u2013 Vi\u1ec7t theo c\u00e1c kh\u00eda c\u1ea1nh kh\u00e1c<\/span>:<\/span><\/h3>\n

Ngo\u00e0i c\u00e1c so s\u00e1nh tr\u00ean, <\/span>ta c\u00f2n c\u00f3 th\u1ec3 khai th\u00e1c EVC \u0111\u1ec3 tr\u1ee3 gi\u00fap vi\u1ec7c so s\u00e1nh Anh-Vi\u1ec7t theo c\u00e1c kh\u00eda c\u1ea1nh kh\u00e1c nh\u01b0: <\/span>s<\/span>o s\u00e1nh t\u1eeb t\u00ecnh th\u00e1i<\/span>,<\/span> ph\u01b0\u01a1ng ti\u1ec7n bi\u1ec3u \u0111\u1ea1t ph\u1ea1m tr\u00f9 th\u1eddi gian, kh\u00f4ng gian gi\u1eefa ti\u1ebfng Anh v\u00e0 ti\u1ebfng Vi\u1ec7t: nh\u1edd v\u00e0o m\u1ed1i <\/span>n\u1ed1i<\/span> t\u1eeb Anh-Vi\u1ec7t v\u00e0 nh\u00e3n ng\u1eef ngh\u0129a c\u1ee7a c\u00e1c t\u1eeb ti\u1ebfng Anh<\/span>,<\/span> ta bi\u1ebft \u0111\u01b0\u1ee3c t\u1eeb n\u00e0o thu\u1ed9c v\u1ec1 ph\u1ea1m tr\u00f9 th\u1eddi gian (TME) v\u00e0 t\u1eeb n\u00e0o thu\u1ed9c v\u1ec1 ph\u1ea1m tr\u00f9 kh\u00f4ng gian (SPA) \u0111\u1ec3 so s\u00e1nh t\u1ef1 \u0111\u1ed9ng. V\u00ed d\u1ee5: v\u1ec1 ph\u1ea1m tr\u00f9 kh\u00f4ng gian, ta th\u1ea5y ng\u01b0\u1eddi Anh d\u00f9ng gi\u1edbi t\u1eeb \u201cin\u201c (trong) khi n\u00f3i \u201c<\/span>in<\/span> the sky\u201d (d\u1ecbch s\u00e1t ngh\u0129a l\u00e0 \u201ctrong tr\u1eddi\u201d), c\u00f2n ng\u01b0\u1eddi Vi\u1ec7t m\u00ecnh s\u1ebd n\u00f3i \u201c<\/span>tr\u00ean<\/span> tr\u1eddi\u201d<\/span>; s<\/span>o s\u00e1nh c\u00e1ch x\u01b0ng h\u00f4 gi\u1eefa ti\u1ebfng Anh v\u00e0 ti\u1ebfng Vi\u1ec7t: c<\/span>\u0169<\/span>ng nh\u1edd v\u00e0o m\u1ed1i <\/span>n\u1ed1i<\/span> t\u1eeb Anh-Vi\u1ec7t v\u00e0 nh\u00e3n t\u1eeb ph\u00e1p, ta bi\u1ebft t\u1eeb n\u00e0o l\u00e0 t\u1eeb ch\u1ec9 v\u1ec1 c\u00e1ch x\u01b0ng h\u00f4 (PRO) \u0111\u1ec3 ta so s\u00e1nh t\u1ef1 \u0111\u1ed9ng.<\/span><\/p>\n

4.3.<\/span>\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>Khai th\u00e1c ph\u1ee5c v\u1ee5 gi\u1ea3ng d\u1ea1y ngo\u1ea1i ng\u1eef<\/span><\/h2>\n

M<\/span>\u1ed9t<\/span> trong nh\u1eefng m\u1ee5c \u0111\u00edch <\/span>ch\u00ednh <\/span>c\u1ee7a vi\u1ec7c x\u00e2y d\u1ef1ng kho ng\u1eef li\u1ec7u song ng\u1eef l\u00e0 \u0111\u1ec3 khai th\u00e1c ph\u1ee5c v\u1ee5 cho vi\u1ec7c gi\u1ea3ng d\u1ea1y <\/span>ngo\u1ea1i ng\u1eef [3], c\u1ee5 th\u1ec3 \u1edf \u0111\u00e2y l\u00e0 d\u1ea1y <\/span>ti\u1ebfng Anh cho ng\u01b0\u1eddi Vi\u1ec7t v\u00e0 ti\u1ebfng Vi\u1ec7t cho ng\u01b0\u1eddi n\u01b0\u1edbc ngo\u00e0i<\/span> th\u00f4ng qua <\/span>c\u00e1c<\/span> so s\u00e1nh tr\u1ef1c quan nh\u01b0:<\/span><\/p>\n

4.3.1.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>Kh\u1ea3o s\u00e1t <\/span>c\u00e1ch d\u00f9ng t\u1eeb qua <\/span>chu\u1ed7i \u0111\u1ed3ng hi\u1ec7n <\/span>(concordance)<\/span>:<\/span><\/h3>\n

M\u1ed9t t\u1eeb c\u00f3 <\/span>th\u1ec3 c\u00f3 <\/span>nhi\u1ec1u ngh\u0129a kh\u00e1c nhau, ngh\u0129a c\u1ee5 th\u1ec3 c\u1ee7a t\u1eeb ph\u1ee5 thu\u1ed9c v\u00e0o ng\u1eef c\u1ea3nh c\u1ee7a t\u1eeb (context). Ch\u00ednh v\u00ec v\u1eady, m\u00e0 khi xem x\u00e9t ngh\u0129a<\/span>\/c\u00e1ch d\u00f9ng<\/span> c\u1ee7a m\u1ed9t t\u1eeb n\u00e0o \u0111\u00f3, ta c\u1ea7n xem x\u00e9t ng\u1eef c\u1ea3nh t\u01b0\u01a1ng \u1ee9ng c\u1ee7a n\u00f3. V\u00ed d\u1ee5: <\/span>c\u00e1ch ch\u1ecdn t\u1eeb ti\u1ebfng Anh ph\u00f9 h\u1ee3p khi d\u1ecbch t\u1eeb \u201cx\u1ea3y ra\u201d<\/span>. Qua quan s\u00e1t trong EVC, ta s\u1ebd t\u1ef1 nghi\u1ec7m ra khi n\u00e0o d\u00f9ng \u201coccur\u201d (nh\u01b0: l\u1ed7i m\u00e1y t\u00ednh), \u201chappen\u201d (tai n\u1ea1n) hay \u201ctake place\u201d (s\u1ef1 ki\u1ec7n)<\/span>.<\/span><\/p>\n

\"paracon\"<\/a><\/p>\n<\/div>\n

H\u00ecnh 4. So s\u00e1nh c\u00e1ch d\u1ecbch t\u1eeb “x\u1ea3y ra” trong ti\u1ebfng Anh.<\/span><\/p>\n

\n

V\u1edbi kho ng\u1eef li\u1ec7u l\u1edbn, \u0111\u01b0\u1ee3c l\u1ea5y m\u1eabu h\u1ee3p l\u00fd, <\/span>h\u1ea7u h\u1ebft c\u00e1c hi\u1ec7n t\u01b0\u1ee3ng ng\u00f4n <\/span>s\u1ebd \u0111\u01b0\u1ee3c ph\u1ea3n \u00e1nh <\/span>ch\u00ednh x\u00e1c, sinh \u0111\u1ed9ng v\u00e0 <\/span>c\u1eadp nh\u1eadt<\/span> h\u01a1n so v\u1edbi t\u1eeb \u0111i\u1ec3n.<\/span> V\u00ed d\u1ee5: <\/span>Trong c\u00e1c t\u1eeb \u0111i\u1ec3n, <\/span>t\u1eeb \u201cfond of\u201d <\/span>\u0111\u01b0\u1ee3c g<\/span>hi l\u00e0 \u201cn\u00e2ng niu, vu\u1ed1t ve\u201d <\/span>(<\/span>mang n\u1ed9i dung<\/span> t\u00edch c\u1ef1c<\/span>), nh\u01b0ng <\/span>qua t<\/span>h\u1ed1ng k\u00ea tr\u00ean ng\u1eef li\u1ec7u th\u1ef1c t\u1ebf BNC\u00a0<\/span>[3] <\/span>(British National Corpus), ng\u01b0\u1eddi ta th\u1ea5y <\/span>h\u01a1n 60% <\/span>t\u1eeb n\u00e0y \u0111\u01b0\u1ee3c<\/span> d\u00f9ng v\u1edbi <\/span>ngh\u0129a ti\u00eau c\u1ef1c (ngh\u0129a l\u00e0 \u201cqu\u1ea5y r\u1ed1i t\u00ecnh d\u1ee5c\u201d !)<\/span>. <\/span><\/p>\n

4.3.2.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>Kh\u1ea3o s\u00e1t chu\u1ed7i ng\u00f4n t\u1eeb<\/span> (co<\/span>llocation<\/span>)<\/span>:<\/span><\/h3>\n

Trong th\u1ef1c t\u1ebf, c\u00f3 m\u1ed9t s\u1ed1 t\u1eeb <\/span>ng\u01b0\u1eddi b\u1ea3n ng\u1eef<\/span> d\u00f9ng<\/span> chung v\u1edbi nhau<\/span>, c<\/span>h\u1eb3ng h\u1ea1n: <\/span>\u0111\u1ecf l\u00f2m\/<\/span>l\u00e8<\/span>, t\u00edm ng\u1eaft<\/span>\/<\/span>l\u00e8<\/span>; g\u00e0 <\/span>tr\u1ed1ng\/<\/span>\u0111\u1ef1c<\/span>, d\u00ea \u0111\u1ef1c<\/span>\/<\/span>tr\u1ed1ng<\/span>; <\/span>s\u00fac mi\u1ec7ng<\/span>\/<\/span>m\u1ed3m<\/span>, \u2026 ;<\/span> big<\/span>\/h<\/span>eavy rain, <\/span>pink<\/span>\/<\/span>ros<\/span>\u00e9<\/span> wine,<\/span> strongly<\/span>\/bitterly disappointed,<\/span> ..<\/span>.<\/span> Ch\u00ednh v\u00ec v\u1eady, v\u1edbi <\/span>EVC, qua vi\u1ec7c <\/span>kh\u1ea3o s\u00e1t <\/span>tr\u1ef1c quan <\/span>c\u00e1c chu\u1ed7i ng\u00f4n t\u1eeb<\/span> n\u00e0y<\/span>,<\/span> s\u1ebd gi\u00fap <\/span>ng\u01b0\u1eddi h\u1ecdc <\/span>ngo\u1ea1i ng\u1eef b<\/span>i\u1ebft c\u00e1ch d\u00f9ng t\u1eeb th\u00edch h\u1ee3p trong ng\u1eef c\u1ea3nh th\u00edch h\u1ee3p<\/span>, bi\u1ebft \u0111\u01b0\u1ee3c t\u00ednh t\u1eeb n\u00e0o s\u1ebd d\u00f9ng v\u1edbi danh t\u1eeb n\u00e0o, \u0111\u1ed9ng t\u1eeb n\u00e0o d\u00f9ng v\u1edbi danh t\u1eeb n\u00e0o,<\/span> tr\u1ea1ng t\u1eeb n\u00e0o \u0111i v\u1edbi \u0111\u1ed9ng t\u1eeb n\u00e0o, <\/span>…<\/span><\/p>\n

4.3.3.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>Chuy\u1ec3n ng\u1eef Anh \u2013 Vi\u1ec7t cho m\u1ed9t s\u1ed1 c\u1ee5m t\u1eeb c\u1ed1 \u0111\u1ecbnh<\/span>:<\/span><\/h3>\n

Trong ng\u00f4n ng\u1eef th\u01b0\u1eddng c\u00f3 c\u00e1c c\u1ee5m t\u1eeb c\u00f3 t\u00ednh th\u00e0nh ng\u1eef cao v\u00e0 khi chuy\u1ec3n ng\u1eef ta kh\u00f4ng th\u1ec3 d\u1ecbch theo ki\u1ec3u tr\u1ef1c ti\u1ebfp t\u1eebng t\u1eeb m\u1ed9t (word-by-word). \u0110\u1ec3 chuy\u1ec3n ng\u1eef \u0111\u01b0\u1ee3c c\u00e1c c\u1ee5m t\u1eeb \u0111\u01b0\u1ee3c ch\u00ednh x\u00e1c, ta c\u00f3 <\/span>th\u1ec3 d\u00f9ng song ng\u1eef \u0111\u1ec3 <\/span>so s\u00e1nh \u0111\u1ed1i chi\u1ebfu tr\u00ean ph\u1ea1m vi r\u1ed9ng v\u00e0<\/span> r\u00fat <\/span>tr\u00edch t\u1ef1 \u0111\u1ed9ng<\/span> danh s\u00e1ch c\u00e1c c\u1ee5m t\u1eeb \u0111\u00f3 \u0111\u1ec3 gi\u1ea3m thi\u1ec3u chi ph\u00ed d\u1ecbch th\u1ee7 c\u00f4ng<\/span> sau n\u00e0y<\/span>.<\/span><\/p>\n

4.4.<\/span>\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>\u1ee8NG D\u1ee4NG<\/span> EVC TRONG C\u00d4NG T\u00c1C D\u1ecaCH THU\u1eacT ANH-VI\u1ec6T<\/span><\/h1>\n

V\u1edbi kho song ng\u1eef Anh-Vi\u1ec7t EVC, ch\u00fang ta c\u00f3 th\u1ec3 s\u1eed d\u1ee5ng m\u1ed9t s\u1ed1 c\u00f4ng ngh\u1ec7 h\u1ed7 tr\u1ee3 d\u1ecbch thu\u1eadt g\u1ea7n \u0111\u00e2y \u0111\u1ec3 gi\u1ea3m thi\u1ec3u \u0111\u00e1ng k\u1ec3 c\u00f4ng s\u1ee9c d\u1ecbch thu\u1eadt v\u0103n b\u1ea3n:<\/span><\/p>\n

4.4.1.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>D\u1ecaCH M\u00c1Y TH\u1ed0NG K\u00ca (STATISTICAL MACHINE TRANSLATION)<\/span><\/h3>\n

Theo c\u00f4ng ngh\u1ec7 d\u1ecbch m\u00e1y th\u1ed1ng k\u00ea (SMT), m\u00e1y s\u1ebd \u201ch\u1ecdc\u201d t\u1ef1 \u0111\u1ed9ng (machine learning) c\u00e1c quy t\u1eafc chuy\u1ec3n ng\u1eef t\u1eeb kho ng\u1eef li\u1ec7u song ng\u1eef l\u1edbn \u0111\u00e3 gi\u00f3ng h\u00e0ng \u1edf m\u1ee9c c\u00e2u (sentence alignment). \u0110\u00e2y l\u00e0 c\u00f4ng ngh\u1ec7 d\u1ecbch t\u1ef1 \u0111\u1ed9ng ph\u1ed5 bi\u1ebfn nh\u1ea5t hi\u1ec7n nay (Google Translate, Bing Translator, … \u0111ang s\u1eed d\u1ee5ng c\u00f4ng ngh\u1ec7 n\u00e0y). V\u1edbi c\u00f4ng ngh\u1ec7 d\u1ecbch t\u1ef1 \u0111\u1ed9ng, ch\u00fang ta xem nh\u01b0 b\u1ea3n d\u1ecbch c\u1ee7a m\u00e1y l\u00e0 b\u1ea3n d\u1ecbch th\u00f4 v\u00e0 ch\u00fang ta ch\u1ec9 ch\u1ec9nh s\u1eeda nh\u1eefng l\u1ed7i sai c\u1ee7a b\u1ea3n d\u1ecbch th\u00f4 \u0111\u00f3 m\u00e0 kh\u00f4ng ph\u1ea3i d\u1ecbch t\u1eeb \u0111\u1ea7u. \u0110i\u1ec1u n\u00e0y gi\u1ea3m \u0111\u00e1ng k\u1ec3 th\u1eddi gian nh\u1eadp li\u1ec7u (c\u00e2u d\u1ecbch) v\u00e0 th\u1eddi gian tra t\u1eeb \u0111i\u1ec3n. N\u1ebfu ng\u1eef li\u1ec7u song ng\u1eef ch\u00fang ta l\u1edbn v\u00e0 c\u00f9ng l\u0129nh v\u1ef1c (kh\u00f4ng ph\u1ea3i l\u00e0 v\u0103n ch\u01b0\u01a1ng) th\u00ec k\u1ebft qu\u1ea3 d\u1ecbch th\u00f4 c\u1ee7a m\u00e1y c\u00e0ng ch\u00ednh x\u00e1c [2].<\/span><\/p>\n

4.4.2.<\/span>\u00a0\u00a0\u00a0\u00a0 <\/span>D\u1ecaCH D\u1ef0A TR\u00caN B\u1ed8 NH\u1eda (TRANSLATION MEMORY)<\/span><\/h3>\n

Trong c\u00e1c c\u00f4ng c\u1ee5 ph\u1ea7n m\u1ec1m d\u1ecbch thu\u1eadt v\u1edbi s\u1ef1 tr\u1ee3 gi\u00fap c\u1ee7a m\u00e1y t\u00ednh (CAT: Computer-Aided Translation), m\u00e1y t\u00ednh s\u1ebd ph\u00e2n t\u00edch t\u1ef1 \u0111\u1ed9ng v\u0103n b\u1ea3n ngu\u1ed3n th\u00e0nh c\u00e1c \u201c\u0111o\u1ea1n\u201d (segment) v\u00e0 l\u01b0u trong b\u1ed9 nh\u1edb d\u01b0\u1edbi d\u1ea1ng song ng\u1eef (t\u1ee9c l\u00e0 c\u00f3 s\u1ef1 li\u00ean k\u1ebft gi\u1eefa \u201c\u0111o\u1ea1n\u201d ngu\u1ed3n v\u1edbi ph\u1ea7n d\u1ecbch t\u01b0\u01a1ng \u1ee9ng). \u201c\u0110o\u1ea1n\u201d \u1edf \u0111\u00e2y l\u00e0 th\u1ec3 l\u00e0 c\u1ee5m t\u1eeb, ng\u1eef, m\u1ec7nh \u0111\u1ec1 hay c\u00e2u. \u0110\u1ec3 khi ta d\u1ecbch v\u0103n b\u1ea3n m\u1edbi, m\u00e1y s\u1ebd t\u1ef1 \u0111\u1ed9ng t\u00ecm ki\u1ebfm trong b\u1ed9 nh\u1edb nh\u1eefng \u201c\u0111o\u1ea1n\u201d n\u00e0o \u0111\u00e3 \u0111\u01b0\u1ee3c d\u1ecbch tr\u01b0\u1edbc \u0111\u00f3 v\u00e0 m\u00e1y s\u1ebd xu\u1ea5t ra c\u00e1c k\u1ebft qu\u1ea3 \u0111\u00e3 d\u1ecbch \u0111\u1ec3 ng\u01b0\u1eddi kh\u00f4ng ph\u1ea3i d\u1ecbch l\u1ea1i. Ng\u01b0\u1eddi ch\u1ec9 d\u1ecbch th\u1ee7 c\u00f4ng c\u00e1c ph\u1ea7n ch\u01b0a \u0111\u01b0\u1ee3c d\u1ecbch tr\u01b0\u1edbc \u0111\u00f3 v\u00e0 m\u00e1y s\u1ebd l\u1ea1i t\u1ef1 c\u1eadp nh\u1eadt ph\u1ea7n d\u1ecbch th\u00eam n\u00e0y \u0111\u1ec3 h\u1ec7 th\u1ed1ng s\u1eed d\u1ee5ng cho l\u1ea7n sau. Vi\u1ec7c ph\u00e2n \u0111o\u1ea1n (segmentation) v\u0103n b\u1ea3n v\u00e0 t\u00ecm ki\u1ebfm c\u00e1c \u0111o\u1ea1n \u0111\u01b0\u1ee3c th\u1ef1c hi\u1ec7n theo c\u00e1c c\u00f4ng c\u1ee5 v\u00e0 gi\u1ea3i thu\u1eadt trong l\u0129nh v\u1ef1c x\u1eed l\u00fd ng\u00f4n ng\u1eef t\u1ef1 nhi\u00ean (NLP: Natural Language Processing). Vi\u1ec7c so s\u00e1nh c\u00e1c \u201c\u0111o\u1ea1n\u201d c\u00f3 th\u1ec3 l\u00e0 so kh\u1edbp ho\u00e0n to\u00e0n ho\u1eb7c so kh\u1edbp m\u1edd (fuzzy) m\u1ed9t c\u00e1ch \u201cth\u00f4ng minh\u201d (khi m\u00e1y x\u00e9t \u0111\u1ebfn c\u00e1c bi\u1ebfn th\u1ec3 v\u1ec1 h\u00ecnh th\u00e1i, ng\u1eef ph\u00e1p v\u00e0 ng\u1eef ngh\u0129a). M\u00e1y s\u1ebd \u01b0u ti\u00ean ch\u1ecdn \u201c\u0111o\u1ea1n\u201d d\u00e0i nh\u1ea5t v\u00e0 ch\u1ecdn ph\u1ea7n d\u1ecbch t\u01b0\u01a1ng \u1ee9ng c\u00f3 x\u00e1c su\u1ea5t cao nh\u1ea5t (v\u00ec c\u00f3 nhi\u1ec1u \u201c\u0111o\u1ea1n\u201d \u0111\u00e3 \u0111\u01b0\u1ee3c d\u1ecbch theo c\u00e1c c\u00e1ch kh\u00e1c nhau t\u00f9y theo ng\u1eef c\u1ea3nh) [3].<\/span><\/p>\n

V\u1edbi c\u00e1ch th\u1ee9c n\u00e0y, n\u1ebfu ch\u00fang ta c\u00f3 kho ng\u1eef li\u1ec7u song ng\u1eef l\u1edbn v\u00e0 c\u00f3 \u0111\u1ed9 t\u01b0\u01a1ng \u0111\u1ed3ng cao (v\u1ec1 t\u1eeb v\u1ef1ng, thu\u1eadt ng\u1eef, c\u1ea5u tr\u00fac, l\u0129nh v\u1ef1c, phong c\u00e1ch) v\u1edbi v\u0103n b\u1ea3n m\u1edbi c\u1ea7n d\u1ecbch (nh\u01b0: b\u1ea3n \u0111\u1ecba h\u00f3a c\u00e1c t\u00e0i li\u1ec7u h\u01b0\u1edbng d\u1eabn s\u1eed d\u1ee5ng, c\u00e1c h\u1ee3p \u0111\u1ed3ng, …), th\u00ec c\u00f4ng s\u1ee9c d\u1ecbch m\u1edbi s\u1ebd \u0111\u01b0\u1ee3c gi\u1ea3m \u0111\u00e1ng k\u1ec3.<\/span><\/p>\n

\"\"<\/p>\n

H\u00ecnh 5<\/span>. <\/span>M\u1ee9c \u0111\u1ed9 ph\u1ed5 bi\u1ebfn c\u1ee7a c\u00e1c c\u00f4ng c\u1ee5 CAT hi\u1ec7n nay.<\/span><\/p>\n

C\u00e1c c\u00f4ng c\u1ee5 n\u00e0y c\u00f2n c\u00f3 th\u1ec3 h\u1ed7 tr\u1ee3 c\u00e1c thao t\u00e1c kh\u00e1c c\u1ee7a ng\u01b0\u1eddi d\u1ecbch, nh\u01b0: gi\u1eef nguy\u00ean \u0111\u1ecbnh d\u1ea1ng t\u1eadp tin ngu\u1ed3n, tra c\u1ee9u thu\u1eadt ng\u1eef, ki\u1ec3m l\u1ed7i ch\u00ednh t\u1ea3, ki\u1ec3m tra t\u00ednh nh\u1ea5t qu\u00e1n vi\u1ec7c d\u1ecbch c\u00e1c thu\u1eadt ng\u1eef, qu\u1ea3n l\u00fd d\u1ef1 \u00e1n d\u1ecbch v\u1edbi nhi\u1ec1u c\u1ed9ng t\u00e1c vi\u00ean t\u1eeb xa, … Tuy nhi\u00ean, c\u00e1c c\u00f4ng c\u1ee5 n\u00e0y hi\u1ec7n nay ch\u1ee7 y\u1ebfu ho\u1ea1t \u0111\u1ed9ng hi\u1ec7u qu\u1ea3 khi ph\u00e2n t\u00edch ti\u1ebfng Anh (v\u00e0 m\u1ed9t s\u1ed1 ti\u1ebfng th\u00f4ng d\u1ee5ng \u1edf Ch\u00e2u \u00c2u), c\u00f2n khi ph\u00e2n t\u00edch ti\u1ebfng Vi\u1ec7t s\u1ebd c\u00f3 nhi\u1ec1u h\u1ea1n ch\u1ebf do \u0111\u1eb7c th\u00f9 ng\u00f4n ng\u1eef ti\u1ebfng Vi\u1ec7t, v\u00ec v\u1eady, c\u1ea7n c\u00f3 s\u1ef1 can thi\u1ec7p (ti\u1ec1n x\u1eed l\u00fd) v\u00e0o ng\u1eef li\u1ec7u tr\u01b0\u1edbc khi s\u1eed d\u1ee5ng ho\u1eb7c ph\u1ea3i s\u1eed d\u1ee5ng ph\u1ea7n m\u1ec1m CAT \u0111\u1eb7c th\u00f9 cho ti\u1ebfng Vi\u1ec7t.<\/span><\/p>\n

Qua ph\u00e2n t\u00edch c\u00e1c \u1ee9ng d\u1ee5ng n\u00f3i tr\u00ean c\u1ee7a song ng\u1eef, ta th\u1ea5y kho EVC c\u1ee7a ch\u00fang ta c\u00f3 th\u1ec3 khai th\u00e1c ph\u1ee5c v\u1ee5 cho c\u1ea3 hai m\u1ee5c \u0111\u00edch: v\u1ec1 l\u00fd thuy\u1ebft (gi\u1ea3ng d\u1ea1y d\u1ecbch thu\u1eadt) l\u1eabn th\u1ef1c h\u00e0nh (c\u00f4ng t\u00e1c d\u1ecbch thu\u1eadt). V\u00ec v\u1eady, ch\u00fang ta c\u1ea7n t\u0103ng c\u01b0\u1eddng s\u1ed1 l\u01b0\u1ee3ng (thu th\u1eadp th\u00eam nhi\u1ec1u v\u0103n b\u1ea3n song ng\u1eef), ch\u1ee7ng lo\u1ea1i (nhi\u1ec1u l\u0129nh v\u1ef1c kh\u00e1c nhau) v\u00e0 x\u1eed l\u00fd s\u00e2u (g\u00e1n nh\u00e3n ng\u00f4n ng\u1eef); \u0111\u1ed3ng th\u1eddi x\u00e2y d\u1ef1ng c\u00e1c c\u00f4ng c\u1ee5 chuy\u1ec3n \u0111\u1ed5i t\u1ef1 \u0111\u1ed9ng \u0111\u1ecbnh d\u1ea1ng t\u1eeb EVC hi\u1ec7n nay sang \u0111\u1ecbnh d\u1ea1ng c\u1ee7a h\u1ec7 SMT\/CAT v\u00e0 ng\u01b0\u1ee3c l\u1ea1i, \u0111\u1ec3 c\u1ea3 2 b\u00ean (sinh vi\u00ean v\u00e0 ng\u01b0\u1eddi d\u1ecbch chuy\u00ean nghi\u1ec7p) h\u1ecdc h\u1ecfi\/k\u1ebf th\u1eeba \u0111\u01b0\u1ee3c to\u00e0n b\u1ed9 tri th\u1ee9c\/kinh nghi\u1ec7m d\u1ecbch thu\u1eadt t\u1eeb h\u00e0ng ng\u00e0n b\u1ea3n d\u1ecbch, h\u00e0ng tri\u1ec7u c\u00e2u \u0111\u00e3 \u0111\u01b0\u1ee3c d\u1ecbch tr\u01b0\u1edbc \u0111\u00e2y.<\/span><\/p>\n

\u00a0<\/span><\/p>\n

5.<\/span>\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>K\u1ebeT LU\u1eacN<\/span><\/h1>\n

Ng\u1eef li\u1ec7u n\u00f3i chung v\u00e0 ng\u1eef li\u1ec7u song ng\u1eef Anh-Vi\u1ec7t n\u00f3i ri\u00eang s\u1ebd gi\u00fap ch\u00fang ta r\u1ea5t nhi\u1ec1u trong v\u00f4 v\u00e0n c\u00e1c \u1ee9ng d\u1ee5ng kh\u00e1c nhau. T\u1eeb l\u0129nh v\u1ef1c Ng\u00f4n ng\u1eef h\u1ecdc so s\u00e1nh \u0111\u1ed1i chi\u1ebfu, cho \u0111\u1ebfn vi\u1ec7c gi\u1ea3ng d\u1ea1y ngo\u1ea1i ng\u1eef n\u00f3i chung, m\u00f4n bi\u00ean d\u1ecbch n\u00f3i ri\u00eang, ng\u1eef li\u1ec7u song ng\u1eef Anh-Vi\u1ec7t s\u1ebd gi\u00fap cho ng\u01b0\u1eddi h\u1ecdc t\u1ef1 \u201cnghi\u1ec7m\u201d ra c\u00e1c quy lu\u1eadt chuy\u1ec3n ng\u1eef m\u00e0 c\u00e1c c\u00e1ch ti\u1ebfp c\u1eadn truy\u1ec1n th\u1ed1ng kh\u00f4ng th\u1ec3 bao qu\u00e1t h\u1ebft \u0111\u01b0\u1ee3c. \u0110\u1eb7c bi\u1ec7t trong c\u00f4ng t\u00e1c d\u1ecbch thu\u1eadt chuy\u00ean nghi\u1ec7p, kho ng\u1eef li\u1ec7u Anh \u2013 Vi\u1ec7t n\u00e0y s\u1ebd l\u00e0m gi\u1ea3m \u0111\u00e1ng k\u1ec3 c\u00f4ng s\u1ee9c d\u1ecbch thu\u1eadt th\u1ee7 c\u00f4ng. N\u1ebfu kho ng\u1eef li\u1ec7u n\u00e0y \u0111\u01b0\u1ee3c ti\u1ebfp t\u1ee5c c\u1eadp nh\u1eadt th\u00ec hi\u1ec7u qu\u1ea3 khai th\u00e1c c\u00e0ng t\u0103ng g\u1ea5p b\u1ed9i. Ngo\u00e0i ra, c\u00e1ch th\u1ee9c n\u00e0y c\u00f3 th\u1ec3 \u00e1p d\u1ee5ng cho c\u00e1c c\u1eb7p ng\u00f4n ng\u1eef kh\u00e1c [4].<\/span><\/p>\n

T\u00c0I LI\u1ec6U THAM KH\u1ea2O<\/span><\/p>\n

[1]. <\/span>Dien Dinh, “Building an Annotated English-Vietnamese parallel Corpus”,<\/span>\u00a0<\/span>MKS: A Journal of Southeast Asian Linguistics and Languages,<\/span> Vol.35, pp.21-36, 2005.<\/span><\/p>\n

[2]. \u0110<\/span>inh<\/span> \u0110i\u1ec1n<\/span>, “<\/span>X\u00e2y d\u1ef1ng v\u00e0 khai th\u00e1c ng\u1eef li\u1ec7u song ng\u1eef Anh-Vi\u1ec7t \u0111i\u1ec7n t\u1eed<\/span>“,<\/span> lu\u1eadn \u00e1n ti\u1ebfn s\u0129 ng\u00f4n ng\u1eef h\u1ecdc so s\u00e1nh<\/span>, \u0110H Khoa h\u1ecdc X\u00e3 h\u1ed9i & Nh\u00e2n v\u0103n, \u0110HQG Tp.HCM, 3\/2005.<\/span><\/p>\n

[3] Lynne Bowker, <\/span>Co<\/span>mputer-Assisted Translation Technology: A Practical Introduction<\/span>, University of Ottawa Press, 2002, ISBN 0-7766-0538-0.<\/span><\/p>\n

[4]\u00a0 <\/span>Phuoc T., Die<\/span>n D. “A N<\/span>ovel <\/span>A<\/span>pproach for <\/span>H<\/span>andling <\/span>U<\/span>nknown <\/span>W<\/span>ord <\/span>P<\/span>roblem<\/span>s<\/span> in <\/span>the <\/span>Chinese-Vietnamese <\/span>M<\/span>achine <\/span>T<\/span>ranslation”, International Journal of Computational Linguistics and Chinese Language Processing (IJCLCLP), Vol.19, No.1, March 2014, pp.1-10, ISSN: 1027-376X<\/span>.<\/span><\/p>\n

PH\u1ea6N M\u1ec0M<\/span><\/p>\n

[1]. <\/span>http:\/\/www.clc.hcmus.edu.vn\/wp-content\/uploads\/resources\/Corpus\/CLC_EVC.zip<\/span><\/a><\/p>\n

[2]. <\/span>http:\/\/www.clc.hcmus.edu.vn\/wp-content\/up<\/span>loads\/resources\/Tools<\/span><\/p>\n

[3]. <\/span>http:\/\/www.statmt.org\/moses\/giza\/GIZA++.html<\/span><\/a><\/p>\n

[4]. <\/span>http:\/\/nlp.stanford.edu\/software\/tagger.shtml<\/span><\/a><\/p>\n

\u00a0—————————————————————<\/span><\/p>\n

(*)<\/strong>\u00a0N\u1ed9i dung b\u00e0i vi\u1ebft n\u00e0y \u0111\u01b0\u1ee3c tr\u00edch t\u1eeb c\u00f4ng tr\u00ecnh:\u00a0<\/b>\u0110inh \u0110i\u1ec1n, L\u00fd Ng\u1ecdc Minh, \u201c\u1ee8ng d\u1ee5ng Ng\u1eef li\u1ec7u Song ng\u1eef Anh-Vi\u1ec7t trong Gi\u1ea3ng d\u1ea1y Ng\u00f4n ng\u1eef\u201d, h\u1ed9i th\u1ea3o\u00a0Li\u00ean ng\u00e0nh NNH \u1ee8ng d\u1ee5ng & Gi\u1ea3ng d\u1ea1y Ng\u00f4n ng\u1eef<\/i>, 11\/2015, Hu\u1ebf, tr.559-567.<\/span><\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

  X\u00e2y d\u1ef1ng v\u00e0 khai th\u00e1c Kho Ng\u1eef li\u1ec7u Song ng\u1eef Anh-Vi\u1ec7t\u00a0(*) 1.\u00a0\u00a0\u00a0\u00a0 T\u1ed4NG QUAN Trong vi\u1ec7c nghi\u00ean c\u1ee9u, gi\u1ea3ng […]<\/p>\n","protected":false},"author":3,"featured_media":0,"parent":226,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":[],"_links":{"self":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/pages\/1506"}],"collection":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1506"}],"version-history":[{"count":11,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/pages\/1506\/revisions"}],"predecessor-version":[{"id":2021,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/pages\/1506\/revisions\/2021"}],"up":[{"embeddable":true,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/pages\/226"}],"wp:attachment":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1506"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}