Machine TranslationChallenges and Language Divergences
Alon LavieLanguage Technologies Institute
Carnegie Mellon University
11-731: Machine TranslationJanuary 12, 2011
January 12, 2011
Major Sources of Translation Problems
• Lexical Differences:– Multiple possible translations for SL word, or
difficulties expressing SL word meaning in a single TL word
• Structural Differences:– Syntax of SL is different than syntax of the TL: word
order, sentence and constituent structure
• Differences in Mappings of Syntax to Semantics:– Meaning in TL is conveyed using a different syntactic
structure than in the SL
• Idioms and Constructions
211-731: Machine Translation
January 12, 2011
Lexical Differences
• SL word has several different meanings, that translate differently into TL– Ex: financial bank vs. river bank
• Lexical Gaps: SL word reflects a unique meaning that cannot be expressed by a single word in TL– Ex: English snub doesn’t have a corresponding verb
in French or German• TL has finer distinctions than SL SL word
should be translated differently in different contexts– Ex: English wall can be German wand (internal),
mauer (external)
311-731: Machine Translation
January 12, 2011
Google at Work…
411-731: Machine Translation
January 12, 2011 511-731: Machine Translation
January 12, 2011 611-731: Machine Translation
January 12, 2011
Lexical Differences
• Lexical gaps: – Examples: these have no direct equivalent in
English:
gratiner(v., French, “to cook with a cheese coating”)
ōtosanrin(n., Japanese, “three-wheeled truck or van”)
711-731: Machine Translation
January 12, 2011
[From Hutchins & Somers]
Lexical Differences
811-731: Machine Translation
January 12, 2011
MT Handling of Lexical Differences
• Direct MT and Syntactic Transfer:– Lexical Transfer stage uses bilingual lexicon– SL word can have multiple translation entries,
possibly augmented with disambiguation features or probabilities
– Lexical Transfer can involve use of limited context (on SL side, TL side, or both)
– Lexical Gaps can partly be addressed via phrasal lexicons
• Semantic Transfer:– Ambiguity of SL word must be resolved during
analysis correct symbolic representation at semantic level
– TL Generation must select appropriate word or structure for correctly conveying the concept in TL
911-731: Machine Translation
January 12, 2011
Structural Differences
• Syntax of SL is different than syntax of the TL: – Word order within constituents:
• English NPs: art adj n the big boy• Hebrew NPs: art n art adj ha yeled ha gadol
– Constituent structure:• English is SVO: Subj Verb Obj I saw the man• Modern Arabic is VSO: Verb Subj Obj
– Different verb syntax:• Verb complexes in English vs. in German I can eat the apple Ich kann den apfel essen
– Case marking and free constituent order• German and other languages that mark case: den apfel esse Ich the(acc) apple eat I(nom)
1011-731: Machine Translation
January 12, 2011 1111-731: Machine Translation
January 12, 2011 1211-731: Machine Translation
January 12, 2011 1311-731: Machine Translation
January 12, 2011
MT Handling of Structural Differences
• Direct MT Approaches:– No explicit treatment: Phrasal Lexicons and sentence
level matches or templates• Syntactic Transfer:
– Structural Transfer Grammars• Trigger rule by matching against syntactic structure on
SL side• Rule specifies how to reorder and re-structure the
syntactic constituents to reflect syntax of TL side
• Semantic Transfer:– SL Semantic Representation abstracts away from SL
syntax to functional roles done during analysis– TL Generation maps semantic structures to correct
TL syntax
1411-731: Machine Translation
January 12, 2011
Syntax-to-Semantics Differences
• Meaning in TL is conveyed using a different syntactic structure than in the SL– Changes in verb and its arguments– Passive constructions– Motion verbs and state verbs– Case creation and case absorption
• Main Distinction from Structural Differences:– Structural differences are mostly independent of
lexical choices and their semantic meaning addressed by transfer rules that are syntactic in nature
– Syntax-to-semantic mapping differences are meaning-specific: require the presence of specific words (and meanings) in the SL
1511-731: Machine Translation
January 12, 2011
Syntax-to-Semantics Differences
• Structure-change example:I like swimming“Ich scwhimme gern”I swim gladly
1611-731: Machine Translation
January 12, 2011 1711-731: Machine Translation
January 12, 2011
Syntax-to-Semantics Differences
• Verb-argument example:Jones likes the film.“Le film plait à Jones.”(lit: “the film pleases to Jones”)
• Use of case roles can eliminate the need for this type of transfer – Jones = Experiencer– film = Theme
1811-731: Machine Translation
January 12, 2011 1911-731: Machine Translation
January 12, 2011 2011-731: Machine Translation
January 12, 2011
Syntax-to-Semantics Differences
• Passive Constructions• Example: French reflexive passives:
Ces livres se lisent facilement*”These books read themselves easily”These books are easily read
2111-731: Machine Translation
January 12, 2011 2211-731: Machine Translation
January 12, 2011 2311-731: Machine Translation
January 12, 2011
Same intention, different syntax
• rigly bitiwgacny my leg hurts• candy wagac fE rigly I have pain in my leg• rigly bitiClimny my leg hurts• fE wagac fE rigly there is pain in my leg• rigly bitinqaH calya my leg bothers on me
Romanization of Arabic from CallHome Egypt.
2411-731: Machine Translation
January 12, 2011
MT Handling of Syntax-to-Semantics Differences
• Direct MT Approaches:– No Explicit treatment: Phrasal Lexicons and sentence
level matches or templates• Syntactic Transfer:
– “Lexicalized” Structural Transfer Grammars• Trigger rule by matching against “lexicalized” syntactic
structure on SL side: lexical and functional features• Rule specifies how to reorder and re-structure the
syntactic constituents to reflect syntax of TL side
• Semantic Transfer:– SL Semantic Representation abstracts away from SL
syntax to functional roles done during analysis– TL Generation maps semantic structures to correct
TL syntax
2511-731: Machine Translation
January 12, 2011
Idioms and Constructions
• Main Distinction: meaning of whole is not directly compositional from meaning of its sub-parts no compositional translation
• Examples:– George is a bull in a china shop– He kicked the bucket– Can you please open the window?
2611-731: Machine Translation
January 12, 2011 2711-731: Machine Translation
January 12, 2011
Formulaic Utterances
• Good night.• tisbaH cala xEr waking up on good
• Romanization of Arabic from CallHome Egypt
2811-731: Machine Translation
January 12, 2011
Constructions
• Identifying speaker intention rather than literal meaning for formulaic and task-oriented sentences.
How about … suggestionWhy don’t you… suggestionCould you tell me… request
info. I was wondering… request info.
2911-731: Machine Translation
January 12, 2011 3011-731: Machine Translation
January 12, 2011
MT Handling of Constructions and Idioms
• Direct MT Approaches:– No Explicit treatment: Phrasal Lexicons and sentence level
matches or templates• Syntactic Transfer:
– No effective treatment– “Highly Lexicalized” Structural Transfer rules can handle
some constructions• Trigger rule by matching against entire construction, including
structure on SL side• Rule specifies how to generate the correct construction on the
TL side• Semantic Transfer:
– Analysis must capture non-compositional representation of the idiom or construction specialized rules
– TL Generation maps construction semantic structures to correct TL syntax and lexical words
3111-731: Machine Translation
January 12, 2011
Take Home Messages• Remember these types of language divergences as you
learn about and apply the various steps in the MT system pipelines of different approaches!
• Ask yourself how capable these various steps and approaches are in addressing these types of divergences!– Can the step/approach handle these divergences?– If so, does it model the divergence at the appropriate level
of abstraction?• Keep these language divergences in mind when you
analyze the errors of the MT system that you have put together and trained!– Are the errors attributable to a particular divergence?– What would be required for the system to address this type
of error?
3211-731: Machine Translation
January 12, 2011
Summary
• Main challenges for current state-of-the-art MT approaches - Coverage and Accuracy:– Acquiring broad-coverage high-accuracy translation
lexicons (for words and phrases)– learning syntactic mappings between languages
from parallel word-aligned data– overcoming syntax-to-semantics differences and
dealing with constructions– Effective Target Language Modeling
3311-731: Machine Translation
January 12, 2011
Homework Assignment #1
3411-731: Machine Translation
January 12, 2011
Questions…
3511-731: Machine Translation
Top Related