{"id":3402,"date":"2023-07-26T23:28:37","date_gmt":"2023-07-26T16:28:37","guid":{"rendered":"https:\/\/www.clc.hcmus.edu.vn\/?p=3402"},"modified":"2023-11-23T14:55:47","modified_gmt":"2023-11-23T07:55:47","slug":"nha-khoa-hoc-viet-dung-ai-dich-chu-nom-sang-chu-quoc-ngu","status":"publish","type":"post","link":"https:\/\/www.clc.hcmus.edu.vn\/?p=3402","title":{"rendered":"Nh\u00e0 khoa h\u1ecdc Vi\u1ec7t d\u00f9ng AI d\u1ecbch ch\u1eef N\u00f4m sang ch\u1eef Qu\u1ed1c ng\u1eef"},"content":{"rendered":"

S\u1eed d\u1ee5ng kho d\u1eef li\u1ec7u h\u00e0ng tr\u0103m tri\u1ec7u t\u1eeb, nh\u00f3m nh\u00e0 khoa h\u1ecdc t\u1ea1i TP HCM x\u00e2y d\u1ef1ng th\u00e0nh c\u00f4ng h\u1ec7 th\u1ed1ng \u1ee9ng d\u1ee5ng tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o (AI) chuy\u1ec3n ch\u1eef N\u00f4m sang ch\u1eef Qu\u1ed1c ng\u1eef.<\/p>\n

\n

Nh\u00f3m nghi\u00ean c\u1ee9u g\u1ed3m 10 gi\u1ea3ng vi\u00ean \u0111\u1ebfn t\u1eeb Khoa C\u00f4ng ngh\u1ec7 Th\u00f4ng tin, \u0110\u1ea1i h\u1ecdc Khoa h\u1ecdc T\u1ef1 nhi\u00ean v\u00e0 B\u1ed9 m\u00f4n H\u00e1n – N\u00f4m, khoa V\u0103n h\u1ecdc, \u0110\u1ea1i h\u1ecdc Khoa h\u1ecdc X\u00e3 h\u1ed9i v\u00e0 Nh\u00e2n v\u0103n (\u0110\u1ea1i h\u1ecdc Qu\u1ed1c gia TP HCM). Nh\u00f3m x\u00e2y d\u1ef1ng h\u1ec7 th\u1ed1ng chuy\u1ec3n ng\u1eef t\u1ef1 \u0111\u1ed9ng (automatic transliteration) t\u1eeb n\u0103m 2020 v\u00e0 hi\u1ec7n \u0111\u00e3 ho\u00e0n th\u00e0nh. Ng\u01b0\u1eddi d\u00f9ng c\u00f3 th\u1ec3 tra c\u1ee9u t\u1ea1i: tools.clc.hcmus.edu.vn.<\/p>\n

\u00dd t\u01b0\u1edfng x\u00e2y d\u1ef1ng h\u1ec7 th\u1ed1ng phi\u00ean d\u1ecbch t\u1ef1 \u0111\u1ed9ng \u0111\u01b0\u1ee3c PGS.TS \u0110inh \u0110i\u1ec1n, Gi\u00e1m \u0111\u1ed1c Trung t\u00e2m ng\u00f4n ng\u1eef h\u1ecdc t\u00ednh to\u00e1n, \u0110\u1ea1i h\u1ecdc Khoa h\u1ecdc T\u1ef1 nhi\u00ean \u1ea5p \u1ee7 t\u1eeb h\u01a1n 20 n\u0103m tr\u01b0\u1edbc. Tuy nhi\u00ean, th\u1eddi \u0111i\u1ec3m \u0111\u00f3 ch\u01b0a c\u00f3 nhi\u1ec1u ngu\u1ed3n d\u1eef li\u1ec7u H\u00e1n – N\u00f4m c\u0169ng nh\u01b0 c\u00e1c m\u00f4 h\u00ecnh m\u00e1y h\u1ecdc ti\u00ean ti\u1ebfn. Nhi\u1ec1u n\u0103m sau, v\u1edbi s\u1ef1 xu\u1ea5t hi\u1ec7n c\u00e1c m\u00f4 h\u00ecnh h\u1ecdc s\u00e2u (deep learning) c\u1ee7a tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o, h\u1ecd m\u1edbi b\u1eaft \u0111\u1ea7u ph\u00e1t tri\u1ec3n m\u00f4 h\u00ecnh phi\u00ean d\u1ecbch t\u1ef1 \u0111\u1ed9ng n\u00e0y.<\/p>\n

\"PGS.TS<\/a>
PGS.TS \u0110inh \u0110i\u1ec1n, Tr\u01b0\u1edfng nh\u00f3m nghi\u00ean c\u1ee9u d\u00f9ng m\u00f4 h\u00ecnh m\u00e1y h\u1ecdc d\u1ecbch ch\u1eef N\u00f4m sang ch\u1eef Qu\u1ed1c ng\u1eef. \u1ea2nh: H\u00e0 An<\/figcaption><\/figure>\n

Nh\u00f3m nghi\u00ean c\u1ee9u thu th\u1eadp ngu\u1ed3n t\u00e0i li\u1ec7u H\u00e1n – N\u00f4m t\u1ea1i c\u00e1c vi\u1ec7n nghi\u00ean c\u1ee9u, th\u01b0 vi\u1ec7n, website, nh\u00e0 khoa h\u1ecdc trong v\u00e0 ngo\u00e0i n\u01b0\u1edbc v\u1edbi kho d\u1eef li\u1ec7u h\u00e0ng tr\u0103m tri\u1ec7u t\u1eeb. D\u1eef li\u1ec7u \u0111\u01b0\u1ee3c s\u1eed d\u1ee5ng m\u00f4 h\u00ecnh lai (hybrid) b\u1eb1ng c\u00e1ch k\u1ebft h\u1ee3p gi\u1eefa m\u00f4 h\u00ecnh m\u00e1y h\u1ecdc d\u1ecbch th\u1ed1ng k\u00ea (SMT: Statistical Machine Translation) v\u00e0 m\u00f4 h\u00ecnh m\u00e1y d\u1ecbch theo m\u1ea1ng n\u01a1ron (NMT: Neural Machine Translation).<\/p>\n

Theo PGS \u0110i\u1ec1n, m\u00f4 h\u00ecnh NMT kh\u1ea3 n\u0103ng d\u1ecbch ng\u00f4n ng\u1eef t\u1ef1 nhi\u00ean t\u1ed1t h\u01a1n, nh\u01b0ng v\u1edbi SMT th\u00ec kh\u1ea3 n\u0103ng chuy\u1ec3n t\u1ef1 ch\u1eef H\u00e1n – N\u00f4m sang ch\u1eef Qu\u1ed1c ng\u1eef l\u1ea1i c\u00f3 \u01b0u th\u1ebf h\u01a1n do kh\u00f4ng c\u00f3 s\u1ef1 thay \u0111\u1ed5i tr\u1eadt t\u1ef1 t\u1eeb nh\u01b0 trong chuy\u1ec3n ng\u1eef th\u01b0\u1eddng g\u1eb7p. V\u00ec v\u1eady, t\u00f9y t\u1eebng tr\u01b0\u1eddng h\u1ee3p, nh\u00f3m s\u1ebd k\u1ebft h\u1ee3p s\u1eed d\u1ee5ng m\u00f4 h\u00ecnh \u0111\u1ec3 cho k\u1ebft qu\u1ea3 t\u1ed1i \u01b0u. V\u1edbi vi\u1ec7c x\u00e2y d\u1ef1ng h\u1ec7 th\u1ed1ng chuy\u1ec3n ng\u1eef ho\u1ea1t \u0111\u1ed9ng tr\u00ean website, kh\u1ea3 n\u0103ng d\u1ecbch c\u1ee7a h\u1ec7 th\u1ed1ng cho \u0111\u1ed9 ch\u00ednh x\u00e1c t\u00f9y theo l\u0129nh v\u1ef1c.<\/p>\n

C\u1ee5 th\u1ec3, v\u1edbi c\u00e1c v\u0103n b\u1ea3n thu\u1ed9c l\u0129nh v\u1ef1c l\u1ecbch s\u1eed, v\u0103n h\u1ecdc, x\u00e3 h\u1ed9i, h\u1ec7 th\u1ed1ng cho k\u1ebft qu\u1ea3 ch\u00ednh x\u00e1c tr\u00ean 90%. C\u00e1c v\u0103n b\u1ea3n v\u1ec1 y h\u1ecdc d\u00e2n t\u1ed9c v\u00e0 c\u00e1c t\u00e0i li\u1ec7u mang t\u00ednh chuy\u00ean ng\u00e0nh, m\u00f4 h\u00ecnh cho \u0111\u1ed9 ch\u00ednh x\u00e1c 70%. Ri\u00eang truy\u1ec7n Ki\u1ec1u m\u00f4 h\u00ecnh c\u00f3 th\u1ec3 d\u1ecbch ch\u00ednh x\u00e1c t\u1edbi 99%.<\/p>\n

\u0110\u1ec3 thu\u1eadn l\u1ee3i trong s\u1eed d\u1ee5ng, nh\u00f3m nghi\u00ean c\u1ee9u \u0111ang ph\u00e1t tri\u1ec3n m\u00f4 h\u00ecnh c\u00f3 th\u1ec3 d\u1ecbch ch\u1eef N\u00f4m tr\u00ean \u1ea3nh ch\u1ee5p. Khi ng\u01b0\u1eddi d\u00f9ng \u0111\u01b0a \u1ea3nh ch\u1ee5p c\u00f3 ch\u1ee9a ch\u1eef N\u00f4m, \u1ee9ng d\u1ee5ng s\u1ebd x\u1eed l\u00fd chuy\u1ec3n th\u00e0nh v\u0103n b\u1ea3n ti\u1ebfng Vi\u1ec7t.<\/p>\n

PGS \u0110i\u1ec1n cho bi\u1ebft, v\u1edbi c\u00e1c v\u0103n b\u1ea3n c\u0169 ch\u1eef b\u1ecb m\u1edd, thi\u1ebfu n\u00e9t, m\u00f4 h\u00ecnh c\u00f3 th\u1ec3 nh\u1eadn d\u1ea1ng sai. Tuy nhi\u00ean, nh\u00f3m \u0111ang nghi\u00ean c\u1ee9u gi\u1ea3i ph\u00e1p c\u00f3 kh\u1ea3 n\u0103ng d\u1ef1 \u0111o\u00e1n ch\u1eef vi\u1ebft d\u1ef1a tr\u00ean n\u00e9t ch\u1eef v\u00e0 ng\u1eef c\u1ea3nh tr\u00ean c\u1ea3 v\u0103n b\u1ea3n \u0111\u1ec3 c\u00f3 th\u1ec3 \u0111o\u00e1n ch\u00ednh x\u00e1c ch\u1eef kh\u00f4ng r\u00f5 r\u00e0ng. Ch\u1ee9c n\u0103ng d\u1ecbch \u1ea3nh \u0111ang trong giai \u0111o\u1ea1n th\u1eed nghi\u1ec7m v\u00e0 ch\u01b0a \u1ee9ng d\u1ee5ng c\u00f4ng khai. K\u1ebft qu\u1ea3 th\u1eed nghi\u1ec7m b\u01b0\u1edbc \u0111\u1ea7u m\u1ed9t s\u1ed1 \u1ea3nh ch\u1ee5p v\u0103n b\u1ea3n ch\u1ea5t l\u01b0\u1ee3ng th\u1ea5p, nh\u01b0ng m\u00f4 h\u00ecnh c\u00f3 th\u1ec3 nh\u1eadn d\u1ea1ng ch\u00ednh x\u00e1c 95%.<\/p>\n

\"Giao<\/a>
Giao di\u1ec7n website chuy\u1ec3n t\u1ef1 ch\u1eef N\u00f4m sang ch\u1eef Qu\u1ed1c ng\u1eef c\u1ee7a nh\u00f3m nghi\u00ean c\u1ee9u.<\/figcaption><\/figure>\n

Theo nh\u00f3m nghi\u00ean c\u1ee9u, \u0111\u00e2y l\u00e0 d\u1ef1 \u00e1n phi l\u1ee3i nhu\u1eadn nh\u1eb1m h\u01b0\u1edbng \u0111\u1ebfn c\u00f4ng c\u1ee5 d\u1ecbch chu\u1ea9n x\u00e1c t\u1eeb ch\u1eef H\u00e1n – N\u00f4m sang ch\u1eef Qu\u1ed1c ng\u1eef. Th\u00f4ng qua d\u1ef1 \u00e1n, c\u1ed9ng \u0111\u1ed3ng c\u00f3 th\u1ec3 \u0111\u00f3ng g\u00f3p ngu\u1ed3n t\u01b0 li\u1ec7u ch\u1eef H\u00e1n – N\u00f4m \u0111\u1ec3 c\u1eadp nh\u1eadt th\u00eam kho ng\u1eef li\u1ec7u hu\u1ea5n luy\u1ec7n m\u00f4 h\u00ecnh phong ph\u00fa h\u01a1n, gi\u00fap ho\u1ea1t \u0111\u1ed9ng ch\u00ednh x\u00e1c h\u01a1n. Th\u00f4ng qua website, c\u00e1c nh\u00e0 nghi\u00ean c\u1ee9u c\u00f3 th\u1ec3 hi\u1ec7u ch\u1ec9nh nh\u1eefng l\u1ed7i nh\u1eadn d\u1ea1ng ch\u1eef H\u00e1n – N\u00f4m sai hay chuy\u1ec3n t\u1ef1 sai, gi\u00fap m\u00e1y ng\u00e0y c\u00e0ng ho\u00e0n thi\u1ec7n h\u01a1n.<\/p>\n

\"Th\u1eed<\/a>
Th\u1eed nghi\u1ec7m kh\u1ea3 n\u0103ng d\u1ecbch c\u1ee7a m\u00f4 h\u00ecnh t\u1eeb \u1ea3nh ch\u1ee5p, \u0111\u1ed9 ch\u00ednh x\u00e1c tr\u00ean 95%. \u1ea2nh: NVCC<\/figcaption><\/figure>\n

TS H\u1ed3 Minh Quang, Tr\u01b0\u1edfng khoa \u0110\u00f4ng Ph\u01b0\u01a1ng H\u1ecdc, \u0110\u1ea1i h\u1ecdc Khoa h\u1ecdc X\u00e3 h\u1ed9i v\u00e0 Nh\u00e2n v\u0103n (\u0110\u1ea1i h\u1ecdc Qu\u1ed1c gia TP HCM) \u0111\u00e1nh gi\u00e1 nghi\u00ean c\u1ee9u c\u00f3 \u00fd ngh\u0129a l\u1edbn trong vi\u1ec7c g\u00ecn gi\u1eef di s\u1ea3n ng\u00f4n ng\u1eef H\u00e1n – N\u00f4m. Tr\u01b0\u1edbc \u0111\u00e2y vi\u1ec7c \u0111\u1ecdc, hi\u1ec3u ch\u1eef N\u00f4m ch\u1ee7 y\u1ebfu l\u00e0 trong gi\u1edbi nghi\u00ean c\u1ee9u. S\u1ea3n ph\u1ea9m c\u1ee7a nh\u00f3m c\u00f3 th\u1ec3 gi\u00fap ng\u01b0\u1eddi d\u00f9ng nh\u1eadn di\u1ec7n, tra c\u1ee9u th\u00f4ng tin sang ch\u1eef Qu\u1ed1c ng\u1eef. \u00d4ng c\u0169ng cho r\u1eb1ng, r\u1ea5t c\u1ea7n s\u1ef1 \u0111\u00f3ng g\u00f3p d\u1eef li\u1ec7u c\u1ee7a c\u1ed9ng \u0111\u1ed3ng \u0111\u1ec3 m\u00f4 h\u00ecnh th\u00f4ng minh, d\u1ecbch chu\u1ea9n x\u00e1c h\u01a1n.<\/p>\n

Ch\u1eef N\u00f4m hi\u1ec7n v\u1eabn c\u00f2n t\u1ed3n t\u1ea1i nhi\u1ec1u trong d\u00e2n gian, nh\u01b0 trong c\u00e1c s\u1eafc phong, gia ph\u1ea3, kh\u1ebf \u01b0\u1edbc, di ch\u00fac, b\u00e0i thu\u1ed1c… C\u00e1c v\u0103n b\u1ea3n n\u00e0y \u0111\u01b0\u1ee3c ghi l\u1ea1i c\u00e1ch \u0111\u00e2y h\u00e0ng tr\u0103m n\u0103m, tr\u00ean c\u00e1c ch\u1ea5t li\u1ec7u ch\u1ea5t l\u01b0\u1ee3ng th\u1ea5p, d\u1ec5 h\u01b0 h\u1ecfng theo th\u1eddi gian n\u1ebfu kh\u00f4ng \u0111\u01b0\u1ee3c b\u1ea3o qu\u1ea3n trong \u0111i\u1ec1u ki\u1ec7n \u0111\u1eb7c bi\u1ec7t. Trong c\u00e1c v\u0103n b\u1ea3n ch\u1eef N\u00f4m, c\u00f3 th\u1ec3 c\u00f3 nhi\u1ec1u th\u00f4ng tin qu\u00fd, nh\u01b0ng ng\u01b0\u1eddi d\u00e2n kh\u00f4ng t\u1ef1 \u0111\u1ecdc hi\u1ec3u \u0111\u01b0\u1ee3c m\u00e0 ph\u1ea3i nh\u1eefng ng\u01b0\u1eddi bi\u1ebft H\u00e1n – N\u00f4m \u0111\u1ec3 phi\u00ean d\u1ecbch qua ch\u1eef Qu\u1ed1c ng\u1eef. Nh\u00f3m nghi\u00ean c\u1ee9u c\u0169ng cho r\u1eb1ng, vi\u1ec7c c\u00f3 c\u00f4ng c\u1ee5 phi\u00ean d\u1ecbch ch\u1eef N\u00f4m sang ch\u1eef Qu\u1ed1c ng\u1eef s\u1ebd gi\u00fap ng\u01b0\u1eddi kh\u00f4ng bi\u1ebft H\u00e1n – N\u00f4m v\u1eabn c\u00f3 th\u1ec3 gi\u1ea3i m\u00e3 th\u00f4ng tin c\u1ee7a nhi\u1ec1u t\u01b0 li\u1ec7u qu\u00fd do t\u1ed5 ti\u00ean h\u1ecd \u0111\u1ec3 l\u1ea1i, trong \u0111\u00f3 c\u00f3 nh\u1eefng b\u00e0i thu\u1ed1c d\u00e2n gian, trong ng\u00e0nh y h\u1ecdc c\u1ed5 truy\u1ec1n \u0111ang l\u01b0u truy\u1ec1n trong nh\u00e2n d\u00e2n.<\/p>\n

H\u00e0 An\u00a0<\/strong><\/p>\n

Ngu\u1ed3n:\u00a0https:\/\/vnexpress.net\/nha-khoa-hoc-viet-dung-ai-dich-chu-nom-sang-chu-quoc-ngu-4621512.html?gidzl=zBM7Jo61uoVxgOP0Bv_XAVglXauvaEfwjwUB6McGkIdXezmPRf6zUE3qsKi_aBKfu_A5G3Nj1bHHAeFXAG<\/a><\/p>\n

 <\/p>\n<\/article>\n

<\/div>\n","protected":false},"excerpt":{"rendered":"

S\u1eed d\u1ee5ng kho d\u1eef li\u1ec7u h\u00e0ng tr\u0103m tri\u1ec7u t\u1eeb, nh\u00f3m nh\u00e0 khoa h\u1ecdc t\u1ea1i TP HCM x\u00e2y d\u1ef1ng th\u00e0nh c\u00f4ng […]<\/p>\n","protected":false},"author":4,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[119,12],"tags":[],"_links":{"self":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/posts\/3402"}],"collection":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=3402"}],"version-history":[{"count":2,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/posts\/3402\/revisions"}],"predecessor-version":[{"id":3407,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=\/wp\/v2\/posts\/3402\/revisions\/3407"}],"wp:attachment":[{"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=3402"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=3402"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.clc.hcmus.edu.vn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=3402"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}