1. Machine Translation Overview Alon Lavie Language Technologies Institute Carnegie Mellon University Open House March 18, 2005

2. LTI Open House2 Machine Translation: History MT started in 1940’s, one of the first conceived application of computers Promising “toy” demonstrations in the 1950’s, failed miserably to scale up to “real” systems AIPAC Report: MT recognized as an extremely difficult, “AI-complete” problem in the early 1960’s MT Revival started in earnest in 1980s (US, Japan) Field dominated by rule-based approaches, requiring 100s of K-years of manual development Economic incentive for developing MT systems for small number of language pairs (mostly European languages)

3. March 18, 2005LTI Open House3 Machine Translation: Where are we today? Age of Internet and Globalization – great demand for MT: –Multiple official languages of UN, EU, Canada, etc. –Documentation dissemination for large manufacturers (Microsoft, IBM, Caterpillar) Economic incentive is still primarily within a small number of language pairs Some fairly good commercial products in the market for these language pairs –Primarily a product of rule-based systems after many years of development Pervasive MT between most language pairs still non- existent and not on the immediate horizon

4. March 18, 2005LTI Open House4 Best Current General-purpose MT PAHO’s Spanam system: Mediante petición recibida por la Comisión Interamericana de Derechos Humanos (en adelante …) el 6 de octubre de 1997, el señor Lino César Oviedo (en adelante …) denunció que la República del Paraguay (en adelante …) violó en su perjuicio los derechos a las garantías judiciales … en su contra. Through petition received by the `Inter-American Commission on Human Rights` (hereinafter …) on 6 October 1997, Mr. Linen César Oviedo (hereinafter “the petitioner”) denounced that the Republic of Paraguay (hereinafter …) violated to his detriment the rights to the judicial guarantees, to the political participation, to // equal protection and to the honor and dignity consecrated in articles 8, 23, 24 and 11, respectively, of the `American Convention on Human Rights` (hereinafter …”), as a consequence of judgments initiated against it.

5. March 18, 2005LTI Open House5 Core Challenges of MT Ambiguity: –Human languages are highly ambiguous, and differently in different languages –Ambiguity at all “levels”: lexical, syntactic, semantic, language-specific constructions and idioms Amount of required knowledge: –At least several 100k words, about as many phrases, plus syntactic knowledge (i.e. translation rules). How do you acquire and construct a knowledge base that big that is (even mostly) correct and consistent?

6. March 18, 2005LTI Open House6 How to Tackle the Core Challenges Manual Labor: 1000s of person-years of human experts developing large word and phrase translation lexicons and translation rules. Example: Systran’s RBMT systems. Lots of Parallel Data: data-driven approaches for finding word and phrase correspondences automatically from large amounts of sentence-aligned parallel texts. Example: Statistical MT systems. Learning Approaches: learn translation rules automatically from small amounts of human translated and word-aligned data. Example: AVENUE’s XFER approach Simplify the Problem: build systems that are limited- domain or constrained in other ways. Examples: CATALYST, NESPOLE!

7. March 18, 2005LTI Open House7 State-of-the-Art in MT What users want: –General purpose (any text) –High quality (human level) –Fully automatic (no user intervention) We can meet any 2 of these 3 goals today, but not all three at once: –FA HQ: Knowledge-Based MT (KBMT) –FA GP: Corpus-Based (Example-Based) MT –GP HQ: Human-in-the-loop (efficiency tool)

8. March 18, 2005LTI Open House8 Types of MT Applications: Assimilation: multiple source languages, uncontrolled style/topic. General purpose MT, no semantic analysis. (GP FA or GP HQ) Dissemination: one source language, controlled style, single topic/domain. Special purpose MT, full semantic analysis. (FA HQ) Communication: Lower quality may be okay, but degraded input, real-time required.

9. March 18, 2005LTI Open House9 Mi chiamo Alon LavieMy name is Alon Lavie Give-information+personal-data (name=alon_lavie) [ s [ vp accusative_pronoun “chiamare” proper_name]] [ s [ np [possessive_pronoun “name”]] [ vp “be” proper_name]] Direct Transfer Interlingua Analysis Generation Approaches to MT: Vaquois MT Triangle

10. March 18, 2005LTI Open House10 Knowledge-based Interlingual MT The “obvious” deep Artificial Intelligence approach: –Analyze the source language into a detailed symbolic representation of its meaning –Generate this meaning in the target language “Interlingua”: one single meaning representation for all languages –Nice in theory, but extremely difficult in practice

11. March 18, 2005LTI Open House11 The Interlingua KBMT approach With interlingua, need only N parsers/ generators instead of N 2 transfer systems: L1 L2 L3 L4 L5 L6 L1 L2 L3 L6 L5 L4 interlingua

12. March 18, 2005LTI Open House12 Statistical MT (SMT) Proposed by IBM in early 1990s: a direct, purely statistical, model for MT Statistical translation models are trained on a sentence-aligned translation corpus Attractive: completely automatic, no manual rules, much reduced manual labor Main drawbacks: –Effective only with huge volumes (several mega- words) of parallel text –Very domain-sensitive –Still viable only for small number of language pairs! Impressive progress in last 3-4 years due to large DARPA funding program (TIDES)

13. March 18, 2005LTI Open House13 EBMT Paradigm New Sentence (Source) Yesterday, 200 delegates met with President Clinton. Matches to Source Found Yesterday, 200 delegates met behind closed doors… Difficulties with President Clinton… Gestern trafen sich 200 Abgeordnete hinter verschlossenen… Schwierigkeiten mit Praesident Clinton… Alignment (Sub-sentential) Translated Sentence (Target) Gestern trafen sich 200 Abgeordnete mit Praesident Clinton. Yesterday, 200 delegates met behind closed doors… Difficulties with President Clinton over… Gestern trafen sich 200 Abgeordnete hinter verschlossenen… Schwierigkeiten mit Praesident Clinton…

14. March 18, 2005LTI Open House14 GEBMT vs. Statistical MT Generalized-EBMT (GEBMT) uses examples at run time, rather than training a parameterized model. Thus: –GEBMT can work with a smaller parallel corpus than Stat MT –Large target language corpus still useful for generating target language model –Much faster to “train” (index examples) than Stat MT; until recently was much faster at run time as well –Generalizes in a different way than Stat MT (whether this is better or worse depends on match between Statistical model and reality): Stat MT can fail on a training sentence, while GEBMT never will GEBMT generalizations based on linguistic knowledge, rather than statistical model design

15. March 18, 2005LTI Open House15 Multi-Engine MT Apply several MT engines to each input; use statistical language modeller to select best combination of outputs. Goal is to combine strengths, and avoid weaknesses. Along all dimensions: domain limits, quality, development time/cost, run-time speed, etc. Used in various projects

16. March 18, 2005LTI Open House16 Speech-to-Speech MT Speech just makes MT (much) more difficult: –Spoken language is messier False starts, filled pauses, repetitions, out-of- vocabulary words Lack of punctuation and explicit sentence boundaries –Current Speech technology is far from perfect Need for speech recognition and synthesis in foreign languages Robustness: MT quality degradation should be proportional to SR quality Tight Integration: rather than separate sequential tasks, can SR + MT be integrated in ways that improves end-to-end performance?

17. March 18, 2005LTI Open House17 MT at the LTI LTI originated as the Center for Machine Translation (CMT) in 1985 MT continues to be a prominent sub-discipline of research with the LTI –More MT faculty than any of the other areas –More MT faculty than anywhere else Active research on all main approaches to MT: Interlingua, Transfer, EBMT, SMT Leader in the area of speech-to-speech MT

18. March 18, 2005LTI Open House18 KBMT: KANT, KANTOO, CATALYST Deep knowledge-based framework, with symbolic interlingua as intermediate representation –Syntactic and semantic analysis into a unambiguous detailed symbolic representation of meaning using unification grammars and transformation mappers –Generation into the target language using unification grammars and transformation mappers First large-scale multi-lingual interlingua-based MT system deployed commercially: –CATALYST at Caterpillar: high quality translation of documentation manuals for heavy equipment Limited domains and controlled English input Minor amounts of post-editing Active follow-on projects Contact Faculty: Eric Nyberg and Teruko Mitamura

19. March 18, 2005LTI Open House19 EBMT Developed originally for the PANGLOSS system in the early 1990s –Translation between English and Spanish Generalized EBMT under development for the past several years Currently one of the two MT approaches developed at CMU for the DARPA/TIDES program –Chinese-to-English, large and very large amounts of sentence-aligned parallel data Active research work on improving alignment and indexing, decoding from a lattice Contact Faculty: Ralf Brown and Jaime Carbonell

20. March 18, 2005LTI Open House20 Statistical MT Word-to-word and phrase-to-phrase translation pairs are acquired automatically from data and assigned probabilities based on a statistical model Extracted and trained from very large amounts of sentence-aligned parallel text –Word alignment algorithms –Phrase detection algorithms –Translation model probability estimation Main approach pursued in CMU systems in the DARPA/TIDES program: –Chinese-to-English and Arabic-to-English Most active work is on phrase detection and on advanced lattice decoding Contact Faculty: Stephan Vogel and Alex Waibel

21. March 18, 2005LTI Open House21 Speech-to-Speech MT Evolution from JANUS/C-STAR systems to NESPOLE!, LingWear, BABYLON –Early 1990s: first prototype system that fully performed sp-to-sp (very limited domain) –Interlingua-based, but with shallow task-oriented representations: “we have single and double rooms available” [give-information+availability] (room-type={single, double}) –Semantic Grammars for analysis and generation –Multiple languages: English, German, French, Italian, Japanese, Korean, and others –Most active work on portable speech translation on small devices: Arabic/English and Thai/English –Contact Faculty: Alan Black, Tanja Schultz and Alex Waibel (also Alon Lavie and Lori Levin)

22. March 18, 2005LTI Open House22 AVENUE: Transfer-based MT Develop new approaches for automatically acquiring syntactic MT transfer rules from small amounts of elicited translated and word-aligned data –Specifically designed to bootstrap MT for languages for which only limited amounts of electronic resources are available (particularly indigenous minority languages) –Use machine learning techniques to generalize transfer rules from specific translated examples –Combine with decoding techniques from SMT for producing the best translation of new input from a lattice of translation segments Languages: Hebrew, Hindi, Mapudungun, Quechua Most active work on designing a typologically comprehensive elicitation corpus, advanced techniques for automatic rule learning, improved decoding, and rule refinement via user interaction Contact Faculty: Alon Lavie, Lori Levin and Jaime Carbonell

23. March 18, 2005LTI Open House23 Transfer with Strong Decoding Learning Module Transfer Rules {PP,4894} ;;Score:0.0470 PP::PP [NP POSTP] -> [PREP NP] ((X2::Y1) (X1::Y2)) Translation Lexicon Run Time Transfer System Lattice Decoder English Language Model Word-to-Word Translation Probabilities Word-aligned elicited data

24. March 18, 2005LTI Open House24 MT for Minority and Indigenous Languages: Challenges Minimal amount of parallel text Possibly competing standards for orthography/spelling Often relatively few trained linguists Access to native informants possible Need to minimize development time and cost

25. March 18, 2005LTI Open House25 Learning Transfer-Rules for Languages with Limited Resources Rationale: –Large bilingual corpora not available –Bilingual native informant(s) can translate and align a small pre-designed elicitation corpus, using elicitation tool –Elicitation corpus designed to be typologically comprehensive and compositional –Transfer-rule engine and new learning approach support acquisition of generalized transfer-rules from the data

26. March 18, 2005LTI Open House26 English-Hindi Example

27. March 18, 2005LTI Open House27 Questions…

28. March 18, 2005LTI Open House28 MEMT chart example Russian leaders signed KBMT (0.8) compact of peace EBMT (0.65) political leaders EBMT (0.9) compact of EBMT (0.7) civilian GLOSS (1.0) tactful DICT (1.0) pact GLOSS (1.0) of peace EBMT (1.0) civil GLOSS (1.0) expedien ts DICT (1.0) bargain DICT (1.0) for DICT (1.0) civil peace EBMT (0.9) political DICT (1.0) Russians DICT (1.0) subscrib e DICT (1.0) pact DICT (1.0) of GLOSS (1.0) quiet DICT (1.0) civilian DICT (1.0) leaders DICT (1.0) politic DICT (1.0) Russian DICT (1.0) sign DICT (1.0) compact DICT (1.0) of DICT (1.0) peace DICT (1.0) civil DICT (1.0) liderespoliticosrusosfirmanpactodepazcivil

29. March 18, 2005LTI Open House29 Why Machine Translation for Minority and Indigenous Languages? Commercial MT economically feasible for only a handful of major languages with large resources (corpora, human developers) Is there hope for MT for languages with limited resources? Benefits include: –Better government access to indigenous communities (Epidemics, crop failures, etc.) –Better indigenous communities participation in information-rich activities (health care, education, government) without giving up their languages. –Language preservation –Civilian and military applications (disaster relief)

30. March 18, 2005LTI Open House30 English-Chinese Example

31. March 18, 2005LTI Open House31 Spanish-Mapudungun Example

32. March 18, 2005LTI Open House32 English-Arabic Example

33. March 18, 2005LTI Open House33 The Elicitation Corpus Translated, aligned by bilingual informant Corpus consists of linguistically diverse constructions Based on elicitation and documentation work of field linguists (e.g. Comrie 1977, Bouquiaux 1992) Organized compositionally: elicit simple structures first, then use them as building blocks Goal: minimize size, maximize linguistic coverage

34. March 18, 2005LTI Open House34 Transfer Rule Formalism Type information Part-of-speech/constituent information Alignments x-side constraints y-side constraints xy-constraints, e.g. ((Y1 AGR) = (X1 AGR)) ; SL: the old man, TL: ha-ish ha-zaqen NP::NP [DET ADJ N] -> [DET N DET ADJ] ( (X1::Y1) (X1::Y3) (X2::Y4) (X3::Y2) ((X1 AGR) = *3-SING) ((X1 DEF = *DEF) ((X3 AGR) = *3-SING) ((X3 COUNT) = +) ((Y1 DEF) = *DEF) ((Y3 DEF) = *DEF) ((Y2 AGR) = *3-SING) ((Y2 GENDER) = (Y4 GENDER)) )

35. March 18, 2005LTI Open House35 Transfer Rule Formalism (II) Value constraints Agreement constraints ;SL: the old man, TL: ha-ish ha-zaqen NP::NP [DET ADJ N] -> [DET N DET ADJ] ( (X1::Y1) (X1::Y3) (X2::Y4) (X3::Y2) ((X1 AGR) = *3-SING) ((X1 DEF = *DEF) ((X3 AGR) = *3-SING) ((X3 COUNT) = +) ((Y1 DEF) = *DEF) ((Y3 DEF) = *DEF) ((Y2 AGR) = *3-SING) ((Y2 GENDER) = (Y4 GENDER)) )

36. March 18, 2005LTI Open House36 The Transfer Engine Analysis Source text is parsed into its grammatical structure. Determines transfer application ordering. Example: 他 看 书。 (he read book) S NP VP N V NP 他 看 书 Transfer A target language tree is created by reordering, insertion, and deletion. S NP VP N V NP he read DET N a book Article “a” is inserted into object NP. Source words translated with transfer lexicon. Generation Target language constraints are checked and final translation produced. E.g. “reads” is chosen over “read” to agree with “he”. Final translation: “He reads a book”

37. March 18, 2005LTI Open House37 Rule Learning - Overview Goal: Acquire Syntactic Transfer Rules Use available knowledge from the source side (grammatical structure) Three steps: 1.Flat Seed Generation: first guesses at transfer rules; flat syntactic structure 2.Compositionality: use previously learned rules to add hierarchical structure 3.Seeded Version Space Learning: refine rules by learning appropriate feature constraints

38. March 18, 2005LTI Open House38 Flat Seed Rule Generation Learning Example: NP Eng: the big apple Heb: ha-tapuax ha-gadol Generated Seed Rule: NP::NP [ART ADJ N]  [ART N ART ADJ] ((X1::Y1) (X1::Y3) (X2::Y4) (X3::Y2))

39. March 18, 2005LTI Open House39 Flat Seed Generation Create a transfer rule that is specific to the sentence pair, but abstracted to the POS level. No syntactic structure. ElementSource SL POS sequencef-structure TL POS sequenceTL dictionary, aligned SL words Type informationcorpus, same on SL and TL Alignmentsinformant x-side constraintsf-structure y-side constraintsTL dictionary, aligned SL words (list of projecting features)

40. March 18, 2005LTI Open House40 Compositionality Initial Flat Rules: S::S [ART ADJ N V ART N]  [ART N ART ADJ V P ART N] ((X1::Y1) (X1::Y3) (X2::Y4) (X3::Y2) (X4::Y5) (X5::Y7) (X6::Y8)) NP::NP [ART ADJ N]  [ART N ART ADJ] ((X1::Y1) (X1::Y3) (X2::Y4) (X3::Y2)) NP::NP [ART N]  [ART N] ((X1::Y1) (X2::Y2)) Generated Compositional Rule: S::S [NP V NP]  [NP V P NP] ((X1::Y1) (X2::Y2) (X3::Y4))

41. March 18, 2005LTI Open House41 Compositionality - Overview Traverse the c-structure of the English sentence, add compositional structure for translatable chunks Adjust constituent sequences, alignments Remove unnecessary constraints, i.e. those that are contained in the lower- level rule

42. March 18, 2005LTI Open House42 Seeded Version Space Learning Input: Rules and their Example Sets S::S [NP V NP]  [NP V P NP] {ex1,ex12,ex17,ex26} ((X1::Y1) (X2::Y2) (X3::Y4)) NP::NP [ART ADJ N]  [ART N ART ADJ] {ex2,ex3,ex13} ((X1::Y1) (X1::Y3) (X2::Y4) (X3::Y2)) NP::NP [ART N]  [ART N] {ex4,ex5,ex6,ex8,ex10,ex11} ((X1::Y1) (X2::Y2)) Output: Rules with Feature Constraints: S::S [NP V NP]  [NP V P NP] ((X1::Y1) (X2::Y2) (X3::Y4) (X1 NUM = X2 NUM) (Y1 NUM = Y2 NUM) (X1 NUM = Y1 NUM))

43. March 18, 2005LTI Open House43 Seeded Version Space Learning: Overview Goal: add appropriate feature constraints to the acquired rules Methodology: –Preserve general structural transfer –Learn specific feature constraints from example set Seed rules are grouped into clusters of similar transfer structure (type, constituent sequences, alignments) Each cluster forms a version space: a partially ordered hypothesis space with a specific and a general boundary The seed rules in a group form the specific boundary of a version space The general boundary is the (implicit) transfer rule with the same type, constituent sequences, and alignments, but no feature constraints

44. March 18, 2005LTI Open House44 Seeded Version Space Learning: Generalization The partial order of the version space: Definition: A transfer rule tr 1 is strictly more general than another transfer rule tr 2 if all f- structures that are satisfied by tr 2 are also satisfied by tr 1. Generalize rules by merging them: –Deletion of constraint –Raising two value constraints to an agreement constraint, e.g. ((x1 num) = *pl), ((x3 num) = *pl)  ((x1 num) = (x3 num))

45. March 18, 2005LTI Open House45 Seeded Version Space Learning NP v det nNP VP … 1.Group seed rules into version spaces as above. 2.Make use of partial order of rules in version space. Partial order is defined via the f-structures satisfying the constraints. 3.Generalize in the space by repeated merging of rules: 1.Deletion of constraint 2.Moving value constraints to agreement constraints, e.g. ((x1 num) = *pl), ((x3 num) = *pl)  ((x1 num) = (x3 num) 4. Check translation power of generalized rules against sentence pairs

46. March 18, 2005LTI Open House46 Seeded Version Space Learning: The Search The Seeded Version Space algorithm itself is the repeated generalization of rules by merging A merge is successful if the set of sentences that can correctly be translated with the merged rule is a superset of the union of sets that can be translated with the unmerged rules, i.e. check power of rule Merge until no more successful merges

47. March 18, 2005LTI Open House47 Seeded VSL: Some Open Issues Three types of constraints: –X-side constrain applicability of rule –Y-side assist in generation –X-Y transfer features from SL to TL Which of the three types improves translation performance? –Use rules without features to populate lattice, decoder will select the best translation… –Learn only X-Y constraints, based on list of universal projecting features Other notions of version-spaces of feature constraints: –Current feature learning is specific to rules that have identical transfer components –Important issue during transfer is to disambiguate among rules that have same SL side but different TL side – can we learn effective constraints for this?

48. March 18, 2005LTI Open House48 Examples of Learned Rules (Hindi-to-English) {NP,14244} ;;Score:0.0429 NP::NP [N] -> [DET N] ( (X1::Y2) ) {NP,14434} ;;Score:0.0040 NP::NP [ADJ CONJ ADJ N] -> [ADJ CONJ ADJ N] ( (X1::Y1) (X2::Y2) (X3::Y3) (X4::Y4) ) {PP,4894} ;;Score:0.0470 PP::PP [NP POSTP] -> [PREP NP] ( (X2::Y1) (X1::Y2) )

49. March 18, 2005LTI Open House49 Manual Transfer Rules: Hindi Example ;; PASSIVE OF SIMPLE PAST (NO AUX) WITH LIGHT VERB ;; passive of 43 (7b) {VP,28} VP::VP : [V V V] -> [Aux V] ( (X1::Y2) ((x1 form) = root) ((x2 type) =c light) ((x2 form) = part) ((x2 aspect) = perf) ((x3 lexwx) = 'jAnA') ((x3 form) = part) ((x3 aspect) = perf) (x0 = x1) ((y1 lex) = be) ((y1 tense) = past) ((y1 agr num) = (x3 agr num)) ((y1 agr pers) = (x3 agr pers)) ((y2 form) = part) )

50. March 18, 2005LTI Open House50 Manual Transfer Rules: Example ; NP1 ke NP2 -> NP2 of NP1 ; Ex: jIvana ke eka aXyAya ; life of (one) chapter ; ==> a chapter of life ; {NP,12} NP::NP : [PP NP1] -> [NP1 PP] ( (X1::Y2) (X2::Y1) ; ((x2 lexwx) = 'kA') ) {NP,13} NP::NP : [NP1] -> [NP1] ( (X1::Y1) ) {PP,12} PP::PP : [NP Postp] -> [Prep NP] ( (X1::Y2) (X2::Y1) ) NP PP NP1 NP P Adj N N1 ke eka aXyAya N jIvana NP NP1 PP Adj N P NP one chapter of N1 N life

51. March 18, 2005LTI Open House51 A Limited Data Scenario for Hindi-to-English Conducted during a DARPA “Surprise Language Exercise” (SLE) in June 2003 Put together a scenario with “miserly” data resources: –Elicited Data corpus: 17589 phrases –Cleaned portion (top 12%) of LDC dictionary: ~2725 Hindi words (23612 translation pairs) –Manually acquired resources during the SLE: 500 manual bigram translations 72 manually written phrase transfer rules 105 manually written postposition rules 48 manually written time expression rules No additional parallel text!!

52. March 18, 2005LTI Open House52 Manual Grammar Development Covers mostly NPs, PPs and VPs (verb complexes) ~70 grammar rules, covering basic and recursive NPs and PPs, verb complexes of main tenses in Hindi (developed in two weeks)

53. March 18, 2005LTI Open House53 Adding a “Strong” Decoder XFER system produces a full lattice of translation fragments, ranging from single words to long phrases or sentences Edges are scored using word-to-word translation probabilities, trained from the limited bilingual data Decoder uses an English LM (70m words) Decoder can also reorder words or phrases (up to 4 positions ahead) For XFER (strong), ONLY edges from basic XFER system are used!

54. March 18, 2005LTI Open House54 Testing Conditions Tested on section of JHU provided data: 258 sentences with four reference translations –SMT system (stand-alone) –EBMT system (stand-alone) –XFER system (naïve decoding) –XFER system with “strong” decoder No grammar rules (baseline) Manually developed grammar rules Automatically learned grammar rules –XFER+SMT with strong decoder (MEMT)

55. March 18, 2005LTI Open House55 Automatic MT Evaluation Metrics Intends to replace or complement human assessment of translation quality of MT produced translation Principle idea: compare how similar is the MT produced translation with human translation(s) of the same input Main metric in use today: IBM’s BLEU –Count n-gram (unigrams, bigrams, trigrams, etc) overlap between the MT output and several reference translations –Calculate a combined n-gram precision score NIST variant of BLEU used for official DARPA evaluations

56. March 18, 2005LTI Open House56 Results on JHU Test Set SystemBLEUM-BLEUNIST EBMT0.0580.1654.22 SMT0.0930.1914.64 XFER (naïve) man grammar 0.0550.1774.46 XFER (strong) no grammar 0.1090.2245.29 XFER (strong) learned grammar 0.1160.2315.37 XFER (strong) man grammar 0.1350.2435.59 XFER+SMT0.1360.2435.65

57. March 18, 2005LTI Open House57 Effect of Reordering in the Decoder

58. March 18, 2005LTI Open House58 Observations and Lessons (I) XFER with strong decoder outperformed SMT even without any grammar rules in the miserly data scenario –SMT Trained on elicited phrases that are very short –SMT has insufficient data to train more discriminative translation probabilities –XFER takes advantage of Morphology Token coverage without morphology: 0.6989 Token coverage with morphology: 0.7892 Manual grammar currently somewhat better than automatically learned grammar –Learned rules did not yet use version-space learning –Large room for improvement on learning rules –Importance of effective well-founded scoring of learned rules

59. March 18, 2005LTI Open House59 Observations and Lessons (II) MEMT (XFER and SMT) based on strong decoder produced best results in the miserly scenario. Reordering within the decoder provided very significant score improvements –Much room for more sophisticated grammar rules –Strong decoder can carry some of the reordering “burden”

60. March 18, 2005LTI Open House60 XFER MT for Hebrew-to-English Two month intensive effort to apply our XFER approach to the development of a Hebrew-to- English MT system Challenges: –No large parallel corpus –Only limited coverage translation lexicon –Morphology: incomplete analyzer available Plan: –Collect available resources, establish methodology for processing Hebrew input –Translate and align Elicitation Corpus –Learn XFER rules –Develop (small) manual XFER grammar as a point of comparison –Evaluate performance on unseen test data using automatic evaluation metrics

61. March 18, 2005LTI Open House61 Hebrew-to-English XFER System First end-to-end integration of system completed yesterday (March 2 nd ) No transfer rules yet, just word-to-word Hebrew-to-English translation No strong decoding yet Amusing Example: office brains the government crack H$BW& in committee the elections the central et the possibility conduct poll crowd about TWKNIT the NSIGH from goat

62. March 18, 2005LTI Open House62 Conclusions Transfer rules (both manual and learned) offer significant contributions that can complement existing data-driven approaches –Also in medium and large data settings? Initial steps to development of a statistically grounded transfer-based MT system with: –Rules that are scored based on a well-founded probability model –Strong and effective decoding that incorporates the most advanced techniques used in SMT decoding Working from the “opposite” end of research on incorporating models of syntax into “standard” SMT systems [Knight et al] Our direction makes sense in the limited data scenario

63. March 18, 2005LTI Open House63 Future Directions Continued work on automatic rule learning (especially Seeded Version Space Learning) Improved leveraging from manual grammar resources, interaction with bilingual speakers Developing a well-founded model for assigning scores (probabilities) to transfer rules Improving the strong decoder to better fit the specific characteristics of the XFER model MEMT with improved –Combination of output from different translation engines with different scorings – strong decoding capabilities

64. March 18, 2005LTI Open House64 Language Modeling for MT Technique stolen from Speech Recognition Try to match the statistics of English Trigram example: “George W. …” Combine quality score with trigram score, to factor in “English-like-ness” Problem: this gives billions of possible overall translations Solution: “beam search”. At each step, throw out all but the “best” possibilities

65. March 18, 2005LTI Open House65 Speech-to-speech translation for eCommerce –CMU, Karlsruhe, IRST, CLIPS, 2 commercial partners Improved limited-domain speech translation Experiment with multimodality and with MEMT EU-side has strict scheduling and deliverables –First test domain: Italian travel agency –Second “showcase”: international Help desk Tied in to CSTAR-III