Language Divergences and Solutions

1
Language Divergences and Solutions

• Advanced Machine Translation Seminar
• Alison Alvarez

2
Overview

• Introduction
• Morphology Primer
• Translation Mismatches
• Types
• Solutions
• Translation Divergences
• Types
• Solutions
• Different MT Systems
• Generation Heavy Machine Translation
• DUSTer

3
Source ? Target

• Languages dont encode the same information in
  the same way
• Makes MT complicated
• Keeps all of us employed

4
Morphology in a Nutshell

• Morphemes are word parts
• Work er
• Iki ta ku na ku na ri ma shi ta
• Types of Morphemes
• Derivational makes new word
• Inflectional adds information to an existing word

5
Morphology in a Nutshell

• Analytic/Isolating
• little or no inflectional morphology, separate
  words
• Vietnamese, Chinese
• I was made to go
• Synthetic
• Lots of inflectional morphology
• Fusional vs. Agglutinating
• Romance Languages, Finnish, Japanese, Mapudungun
• Ika (to go) se (to make/let) rare (passive) ta
  (past tense)
• He need s (3rd person singular) it.

6
Translation Differences

• Types
• Translation Mismatches
• Different information from source to target
• Translation Divergences
• Same information from source to target, but the
  meaning is distributed differently in each
  language

7
Translation Mismatches

• the information that is conveyed is different
  in the source and target languages
• Types
• Lexical level
• Typological level

8
Lexical Mismatches

• A lexical item in one language may have more
  distinctions than in another

Brother
? otouto Younger Brother
??? Ani-san Older Brother
9
Typological Mismatches

• Mismatch between languages with different levels
  of grammaticalization
• One language may be more structurally complex
• Source marking, Obligatory Subject

10
Typological Mismatches

• Source Quechua vs. English
• (they say) s/he was singing --gt takisharansi
• taki (sing) sha (progressive) ra (past) n
  (3rd sg) si (reportative)
• Obligatory Arguments English vs. Japanese
• Kusuri wo Nonda --gt (I, you, etc.) took medicine.
• Makasemasu! --gt(Ill) leave (it) to (you)

11
Translation Mismatch Solutions

• More information --gt Less information (easy)
• Less information --gt More information (hard)
• Context clues
• Language Models
• Generalization
• Formal representations

12
Translation Divergences

• the same information is conveyed in source and
  target texts
• Divergences are quite common
• Occurs in about 1 out of every three sentences in
  the TREC El Norte Newspaper corpus
  (Spanish-English)
• Sentences can have multiple kinds of divergences

13
Translation Divergence Types

• Categorial Divergence
• Conflational Divergence
• Structural Divergence
• Head Swapping Divergence
• Thematic Divergence

14
Categorial Divergence

• Translation that uses different parts of speech
• Tener hambre (have hunger) --gt be hungry
• Noun --gt adjective

15
Conflational Divergence

• The translation of two words using a single word
  that combines their meaning
• Can also be called a lexical gap
• X stab Z --gt X dar puñaladas a Z (X give stabs to
  Z)
• glastuinbouw --gt cultivation under glass

16
Structural Divergence

• A difference in the realization of incorporated
  arguments
• PP to Object
• X entrar en Y (X enter in Y) --gt X enter Y
• X ask for a referendum --gt X pedir un referendum
  (ask-for a referendum)

17
Head Swapping Divergence

• Involves the demotion of a head verb and the
  promotion of a modifier verb to head position

S NP VP N V PP I ran into the room.
S NP VP N V PP VP Yo entro en el cuarto
corriendo
18
Thematic Divergence

• This divergence occurs when sentence arguments
  switch argument roles from one language to
  another
• X gustar a Y (X please to Y) --gt Y like X

19
Divergence Solutions and Statistical/EBMT Systems

• Not really addressed explicitly in SMT
• Covered in EBMT only if it is covered extensively
  in the data

20
Divergence Solutions and Transfer Systems

• Hand-written transfer rules
• Automatic extraction of transfer rules from
  bi-texts
• Problematic with multiple divergences

21
Divergence Solutions and Interlingua Systems

• Melcuks Deep Syntactic Structure
• Jackendoffs Lexical Semantic Structure
• Both require explicit symmetric knowledge from
  both source and target language
• Expensive

22
Divergence Solutions and Interlingua Systems
John swam across a river
event CAUSE JOHN event GO JOHN path ACROSS
JOHN position AT JOHN RIVER manner
SWIMINGLY
Juan cruza el río nadando
23
Generation-Heavy MT

• Built to address language divergences
• Designed for source-poor/target-rich translation
• Non-Interlingual
• Non-Transfer
• Uses symbolic overgeneration to account for
  different translation divergences

24
Generation-Heavy MT

• Source language
• syntactic parser
• translation lexicon
• Target language
• lexical semantics, categorial variations
  subcategorization frames for overgeneration
• Statistical language model

25
GHMT System
26
Analysis Stage

• Independent of Target Language
• Creates a deep syntactic dependency
• Only argument structure, top-level conceptual
  nodes thematic-role information
• Should normalize over syntactic morphological
  phenomena

27
Translation Stage

• Converts SL lexemes to TL lexemes
• Maintains dependency structure

28
Analysis/Translation Stage
GIVE (v) cause go
I agent
STAB (n) theme
JOHN goal
29
Generation Stage

• Lexical Structural Selection
• Conversion to a thematic dependency
• Uses syntactic-thematic linking map
• loose linking
• Structural expansion
• Addresses conflation head-swapped divergences
• Turn thematic dependency to TL syntactic
  dependency
• Addresses categorial divergence

30
Generation Stage Structural Expansion
31
Generation Stage

• Linearization Step
• Creates a word lattice to encode different
  possible realizations
• Implemented using oxyGen engine
• Sentences ranked extracted
• Nitrogens statistical extractor

32
Generation Stage
33
GHMT Results

• 4 of 5 Spanish-English divergences can be
  generated using structural expansion categorial
  variations
• The remaining 1 out of 5 needed more world
  knowledge or idiom handling
• SL syntactic parser can still be hard to come by

34
Divergences and DUSTer

• Helps to overcome divergences for word alignment
  improve coder agreement
• Changes an English sentence structure to resemble
  another language
• More accurate alignment and projection of
  dependency trees without training on dependency
  tree data

35
DUSTer

• Motivation for the development of automatic
  correction of divergences
• Every Language Pair has translation divergences
  that are easy to recognize
• Knowing what they are and how to accommodate
  them provides the basis for refined word level
  alignment
• Refined word-level alignment results in
  improved projection of structural information
  from English to another language

36
DUSTer
37
DUSTer

• Bi-text parsed on English side only
• Linguistically Motivated common search terms
• Conducted on Spanish Arabic (and later Chinese
  Hindi)
• Uses all of the divergences mentioned before,
  plus a light verb divergence
• Try ? put to trying ? poner a prueba

38
DUSTer Rule Development Methods

• Identify canonical transformations for each
  divergence type
• Categorize English sentences into divergence type
  or none
• Apply appropriate transformations
• Humans align E ? E ? foreign language

39
DUSTer Rules

• "kill" gt "LightVB kill(N)" (LightVB light
  verb)
• Presumably, this will work for "kill" gt "give
  death to
• "borrow" gt "take lent (thing) to
• "hurt" gt "make harm to
• "fear" gt "have fear of
• "desire" gt "have interest in
• "rest" gt "have repose on
• "envy" gt "have envy of
• type1.B.X English2 1 3 Spanish2 1 3 4 5
• Verblt1,i,CatVarV_Ngt Nounlt2,j,Subjgt
  Nounlt3,k,Objgt lt--gt
• LightVBlt1,Verbgt Nounlt2,j,Subjgt
  Nounlt3,i,Objgt Obliquelt4,Pred,Prepgt
  Nounlt5,k,PObjgt

40
DUSTer Results
41
Conclusion

• Divergences are common
• They are not handled well by most MT systems
• GHMT can account for divergences, but still needs
  development
• DUSTer can handle divergences through structure
  transformations, but requires a great deal of
  linguistic knowledge

42
The End

• Questions?

43
References

• Dorr, Bonnie J., "Machine Translation
  Divergences A Formal Description and Proposed
  Solution," Computational Linguistics, 204, pp.
  597--633, 1994.
• Dorr, Bonnie J. and Nizar Habash, "Interlingua
  Approximation A Generation-Heavy Approach", In
  Proceedings of Workshop on Interlingua
  Reliability, Fifth Conference of the Association
  for Machine Translation in the Americas,
  AMTA-2002,Tiburon, CA, pp. 1--6, 2002
• Dorr, Bonnie J., Clare R. Voss, Eric Peterson,
  and Michael Kiker, "Concept Based Lexical
  Selection," Proceedings of the AAAI-94 fall
  symposium on Knowledge Representation for Natural
  Language Processing in Implemented Systems, New
  Orleans, LA, pp. 21--30, 1994.
• Dorr, Bonnie J., Lisa Pearl, Rebecca Hwa, and
  Nizar Habash, "DUSTer A Method for Unraveling
  Cross-Language Divergences for Statistical
  Word-Level Alignment," Proceedings of the Fifth
  Conference of the Association for Machine
  Translation in the Americas, AMTA-2002,Tiburon,
  CA, pp. 31--43, 2002.
• Habash, Nizar and Bonnie J. Dorr, "Handling
  Translation Divergences Combining Statistical
  and Symbolic Techniques in Generation-Heavy
  Machine Translation", In Proceedings of the Fifth
  Conference of the Association for Machine
  Translation in the Americas, AMTA-2002,Tiburon,
  CA, pp. 84--93, 2002.
• Haspelmath, Martin. Understanding Morphology.
  Oxford Univeristy Press, 2002.
• Kameyama, Megumi and Ryo Ochitani, Stanley
  Peters Resolving Translation Mismatches With
  Information Flow Annual Meeting of the
  Assocation of Computational Linguistics, 1991

44
Other Divergences

• Idioms
• Aspectual Divergences
• Knowledge outside of Lexical Semantics