Invited Lecture Introduction to Natural Language Processing Fall

• Currently, Google offers translations between the following languages: Arabic Bulgarian Catalan Chinese

Thank you for your attention! Questions?

Road Map • Multilingual Challenges for MT • MT Approaches • MT Evaluation

Multilingual Challenges • Orthographic Variations – Ambiguous spelling • ﺍﻷﻻ ﺍﺷﺍﺭ ﻛﺘﺐ ﺍﻻﻭﻻﺩ ﺍﺷﻌﺎﺭﺍ

Multilingual Challenges Morphological Variations • Affixation vs. Root+Pattern written killed done write kill do

Translation Divergences conflation ﻟﺴﺖ am ﻫﻨﺎ ﻟﺴﺖ ﻫﻨﺎ I-am-not here I not suis here

Translation Divergences head swap and categorial English John swam across the river quickly Spanish

MT Approaches MT Pyramid Source meaning Target meaning Source syntax Source word Analysis Target

MT Approaches Gisting Example Sobre la base de dichas experiencias se estableció en 1988

MT Approaches MT Pyramid Source meaning Source syntax Source word Analysis Target meaning Transfer

MT Approaches Transfer Example • Transfer Lexicon – Map SL structure to TL structure

MT Approaches MT Pyramid Source meaning Source syntax Source word Analysis Interlingua Transfer Gisting

MT Approaches Interlingua Example: Lexical Conceptual Structure (Dorr, 1993)

MT Approaches MT Pyramid Source meaning Interlingual Lexicons Source syntax Source word Analysis Transfer

MT Approaches Statistical vs. Rule-based Source meaning Source syntax Source word Analysis Target meaning

Statistical MT Noisy Channel Model Portions from http: //www. clsp. jhu. edu/ws 03/preworkshop/lecture_yamada. pdf

Slide based on Kevin Knight’s http: //www. sims. berkeley. edu/courses/is 290 -2/f 04/lectures/mt-lecture. ppt

Statistical MT IBM Model (Word-based Model) http: //www. clsp. jhu. edu/ws 03/preworkshop/lecture_yamada. pdf

Slide courtesy of Kevin Knight http: //www. sims. berkeley. edu/courses/is 290 -2/f 04/lectures/mt-lecture. ppt

MT Approaches Statistical vs. Rule-based vs. Hybrid Source meaning Source syntax Source word Analysis

MT Approaches Practical Considerations • Resources Availability – Parsers and Generators • Input/Output compatability

MT Evaluation • More art than science • Wide range of Metrics/Techniques – interface,

Human-based Evaluation Example Accuracy Criteria

Human-based Evaluation Example Fluency Criteria

Automatic Evaluation Example Bleu Metric (Papineni et al 2001) • Bleu – – –

Automatic Evaluation Example Bleu Metric Test Sentence Gold Standard References colorless green ideas sleep

Automatic Evaluation Example METEOR (Lavie and Agrawal 2007) • Metric for Evaluation of Translation

Metrics MATR Workshop • Workshop in AMTA conference 2008 – Association for Machine Translation

Automatic Evaluation Example SEPIA (Habash and El. Kholy 2008) • A syntactically-aware evaluation metric

Interested in MT? ? • Contact me (habash@cs. columbia. edu) • Research courses, projects

Slides: 44

Download presentation

Invited Lecture Introduction to Natural Language Processing Fall 2008 Machine Translation: Challenges and Approaches Nizar Habash Associate Research Scientist Center for Computational Learning Systems Columbia University

• Currently, Google offers translations between the following languages: Arabic Bulgarian Catalan Chinese Croatian Czech Danish Dutch Filipino Finnish French German Greek Hebrew Hindi Indonesian Italian Japanese Korean Latvian Lithuanian Norwegian Polish Portuguese Romanian Russian Serbian Slovak Slovenian Spanish Swedish Ukrainian Vietnamese

Thank you for your attention! Questions?

“BBC found similar support”!!!

Road Map • Multilingual Challenges for MT • MT Approaches • MT Evaluation

Multilingual Challenges • Orthographic Variations – Ambiguous spelling • ﺍﻷﻻ ﺍﺷﺍﺭ ﻛﺘﺐ ﺍﻻﻭﻻﺩ ﺍﺷﻌﺎﺭﺍ – Ambiguous word boundaries • • Lexical Ambiguity – Bank ( ﺑﻨﻚ financial) vs. ( ﺿﻔﺔ river) – Eat essen (human) vs. fressen (animal)

Multilingual Challenges Morphological Variations • Affixation vs. Root+Pattern written killed done write kill do ﻛﺘﺐ ﻗﺘﻞ ﻓﻌﻞ ﻣﻜﺘﻮﺏ ﻣﻘﺘﻮﻝ ﻣﻔﻌﻮﻝ • Tokenization conj noun article plural And the cars ﻭﺍﻟﺴﻴﺎﺭﺍﺕ Et les voitures and the cars w Al Sy. Ar. At et le voitures

Translation Divergences conflation ﻟﺴﺖ am ﻫﻨﺎ ﻟﺴﺖ ﻫﻨﺎ I-am-not here I not suis here I am not here Je ne pas ici Je ne suis pas ici I not am not here

Translation Divergences head swap and categorial English John swam across the river quickly Spanish Juan cruzó rapidamente el río nadando Gloss: John crossed fast the river swimming Arabic ﺍﺳﺮﻉ ﺟﻮﻥ ﻋﺒﻮﺭ ﺍﻟﻨﻬﺮ ﺳﺒﺎﺣﺔ Gloss: sped john crossing the-river swimming Chinese 约翰快速地游过这条河 Gloss: John quickly (DE) swam cross the (Quantifier) river Russian Джон быстро переплыл реку Gloss: John quickly cross-swam river

Road Map • Multilingual Challenges for MT • MT Approaches • MT Evaluation

MT Approaches MT Pyramid Source meaning Target meaning Source syntax Source word Analysis Target syntax Gisting Target word Generation

MT Approaches Gisting Example Sobre la base de dichas experiencias se estableció en 1988 una metodología. Envelope her basis out speak experiences them settle at 1988 one methodology. On the basis of these experiences, a methodology was arrived at in 1988.

MT Approaches MT Pyramid Source meaning Source syntax Source word Analysis Target meaning Transfer Gisting Target syntax Target word Generation

MT Approaches Transfer Example • Transfer Lexicon – Map SL structure to TL structure poner : subj X : obj mantequilla butter : mod en : obj : subj X : obj Y Y X puso mantequilla en Y X buttered Y

MT Approaches MT Pyramid Source meaning Source syntax Source word Analysis Interlingua Transfer Gisting Target meaning Target syntax Target word Generation

MT Approaches Interlingua Example: Lexical Conceptual Structure (Dorr, 1993)

MT Approaches MT Pyramid Source meaning Source syntax Source word Analysis Interlingua Transfer Gisting Target meaning Target syntax Target word Generation

MT Approaches MT Pyramid Source meaning Interlingual Lexicons Source syntax Source word Analysis Transfer Lexicons Target meaning Target syntax Dictionaries/Parallel Corpora Target word Generation

MT Approaches Statistical vs. Rule-based Source meaning Source syntax Source word Analysis Target meaning Target syntax Target word Generation

Statistical MT Noisy Channel Model Portions from http: //www. clsp. jhu. edu/ws 03/preworkshop/lecture_yamada. pdf

Slide based on Kevin Knight’s http: //www. sims. berkeley. edu/courses/is 290 -2/f 04/lectures/mt-lecture. ppt Statistical MT Automatic Word Alignment • GIZA++ – A statistical machine translation toolkit used to train word alignments. – Uses Expectation-Maximization with various constraints to bootstrap alignments Maria Mary did not slap the green witch no dio una bofetada a la bruja verde

Statistical MT IBM Model (Word-based Model) http: //www. clsp. jhu. edu/ws 03/preworkshop/lecture_yamada. pdf

Slide courtesy of Kevin Knight http: //www. sims. berkeley. edu/courses/is 290 -2/f 04/lectures/mt-lecture. ppt Phrase-Based Statistical MT Morgen fliege ich Tomorrow I will fly nach Kanada to the conference zur Konferenz In Canada • Foreign input segmented in to phrases – “phrase” is any sequence of words • Each phrase is probabilistically translated into English – P(to the conference | zur Konferenz) – P(into the meeting | zur Konferenz) • Phrases are probabilistically re-ordered See [Koehn et al, 2003] for an intro. This is state-of-the-art!

Slide courtesy of Kevin Knight http: //www. sims. berkeley. edu/courses/is 290 -2/f 04/lectures/mt-lecture. ppt Word Alignment Induced Phrases Maria no dió una bofetada a la bruja verde Mary did not slap the green witch (Maria, Mary) (no, did not) (slap, dió una bofetada) (la, the) (bruja, witch) (verde, green) (a la, the) (dió una bofetada a, slap the) (Maria no, Mary did not) (no dió una bofetada, did not slap), (dió una bofetada a la, slap the) (bruja verde, green witch)

Slide courtesy of Kevin Knight http: //www. sims. berkeley. edu/courses/is 290 -2/f 04/lectures/mt-lecture. ppt Advantages of Phrase-Based SMT • Many-to-many mappings can handle noncompositional phrases • Local context is very useful for disambiguating – “Interest rate” … – “Interest in” … • The more data, the longer the learned phrases – Sometimes whole sentences

MT Approaches Statistical vs. Rule-based vs. Hybrid Source meaning Source syntax Source word Analysis Target meaning Target syntax Target word Generation

MT Approaches Practical Considerations • Resources Availability – Parsers and Generators • Input/Output compatability – Translation Lexicons • Word-based vs. Transfer/Interlingua – Parallel Corpora • Domain of interest • Bigger is better • Time Availability – Statistical training, resource building

Road Map • Multilingual Challenges for MT • MT Approaches • MT Evaluation

MT Evaluation • More art than science • Wide range of Metrics/Techniques – interface, …, scalability, …, faithfulness, . . . space/time complexity, … etc. • Automatic vs. Human-based – Dumb Machines vs. Slow Humans

Human-based Evaluation Example Accuracy Criteria

Human-based Evaluation Example Fluency Criteria

Automatic Evaluation Example Bleu Metric (Papineni et al 2001) • Bleu – – – Bi. Lingual Evaluation Understudy Modified n-gram precision with length penalty Quick, inexpensive and language independent Correlates highly with human evaluation Bias against synonyms and inflectional variations

Automatic Evaluation Example Bleu Metric Test Sentence Gold Standard References colorless green ideas sleep furiously all dull jade ideas sleep irately drab emerald concepts sleep furiously colorless immature thoughts nap angrily Unigram precision = 4 / 5 = 0. 8 Bigram precision = 2 / 4 = 0. 5 Bleu Score = (a 1 a 2 …an)1/n = (0. 8 ╳ 0. 5)½ = 0. 6325 63. 25

Automatic Evaluation Example METEOR (Lavie and Agrawal 2007) • Metric for Evaluation of Translation with Explicit word Ordering • Extended Matching between translation and reference – Porter stems, word. Net synsets • Unigram Precision, Recall, parameterized F-measure • Reordering Penalty • Parameters can be tuned to optimize correlation with human judgments • Not biased against “non-statistical” MT systems

Metrics MATR Workshop • Workshop in AMTA conference 2008 – Association for Machine Translation in the Americas • Evaluating evaluation metrics • Compared 39 metrics – 7 baselines and 32 new metrics – Various measures of correlation with human judgment – Different conditions: text genre, source language, number of references, etc.

Automatic Evaluation Example SEPIA (Habash and El. Kholy 2008) • A syntactically-aware evaluation metric – (Liu and Gildea, 2005; Owczarzak et al. , 2007; Giménez and Màrquez, 2007) • Uses dependency representation – MICA parser (Nasr & Rambow 2006) – 77% of all structural bigrams are surface n-grams of size 2, 3, 4 • Includes dependency surface span as a factor in score – long-distance dependencies should receive a greater weight than short distance dependencies • Higher degree of grammaticality?

Interested in MT? ? • Contact me (habash@cs. columbia. edu) • Research courses, projects • Languages of interest: – English, Arabic, Hebrew, Chinese, Urdu, Spanish, Russian, …. • Topics – Statistical, Hybrid MT • Phrase-based MT with linguistic extensions • Component improvements or full-system improvements – MT Evaluation – Multilingual computing

Thank You