SlideShare a Scribd company logo

Identification of Fertile Translations in Comparable Corpora: a Morpho-Compositional Approach

Estelle Delpech, profile picture
Estelle Delpech

The document discusses a morpho-compositional approach for generating domain-specific bilingual lexicons using comparable corpora when parallel data is unavailable. It outlines the challenges of translation variation and proposes a compositional translation framework to address morphological and lexical variations. The paper emphasizes the importance of understanding domain-specific translations to improve translation accuracy in technical and scientific contexts.

TechnologyBusiness
Related topics:
Machine Translation
1 of 141
Download to read offline
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
Identification of Fertile Translations in Comparable Corpora a Morpho-Compositional Approach Estelle Delpech1 , B´atrice Daille1 , Emmanuel Morin1 , Claire e Lemaire2,3 1 LINA, 2 GREMUTS, Universit´ de Grenoble Universit´ de Nantes e e 3 Lingua et Machina AMTA’12 10/31/12 San Diego, CA
Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context 1 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context Research partly funded by Computer-Aided Translation company 1 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context Research partly funded by Computer-Aided Translation company Goal: generate domain-specific bilingual lexicons when no parallel data is available 1 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context Research partly funded by Computer-Aided Translation company Goal: generate domain-specific bilingual lexicons when no parallel data is available Available data: general language bilingual dictionary domain-specific comparable corpora 1 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs 2 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: 2 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts 2 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts mixed text types : technical, scientific, lay science... 2 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts mixed text types : technical, scientific, lay science... ⇒ do not expect parallelism in source ↔ target structures 2 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts mixed text types : technical, scientific, lay science... ⇒ do not expect parallelism in source ↔ target structures ⇒ need to deal with variation in translation 2 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation 3 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e 3 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons 3 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance 3 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance ⇒ use of synonyms, thesaurus 3 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance ⇒ use of synonyms, thesaurus Fertility: bi-dimensional → deux dimensions ’two dimensions’ 3 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance ⇒ use of synonyms, thesaurus Fertility: bi-dimensional → deux dimensions ’two dimensions’ ⇒ scarcely adressed 3 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility 4 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility Definition target term has more words than the source term 4 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility Definition target term has more words than the source term semantic fertility target term has more morphemes than the source term voie de glace ’route of ice’ → ice climbing route aquarelle (not decomposable) → water color 4 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility Definition target term has more words than the source term semantic fertility target term has more morphemes than the source term voie de glace ’route of ice’ → ice climbing route aquarelle (not decomposable) → water color surface fertility target and source terms have the same number of morphemes bi-dimensional → deux dimensions ’two dimensions’ 4 / 28
Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation Principle of compositionality “the meaning of the whole is a function of the meaning of the parts” [Keenan and Faltz, 1985, 24-25] 5 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation Principle of compositionality “the meaning of the whole is a function of the meaning of the parts” [Keenan and Faltz, 1985, 24-25] Definition of compositional translation The translation of the whole is a function of the translation of the parts 5 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 6 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 6 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 2 Translation {cyto, toxic} → {cyto, toxique} 6 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 2 Translation {cyto, toxic} → {cyto, toxique} 3 Recomposition {cyto, toxique} → {cytotoxique, toxiquecyto} 6 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 2 Translation {cyto, toxic} → {cyto, toxique} 3 Recomposition {cyto, toxique} → {cytotoxique, toxiquecyto} 4 Selection {cytotoxique, toxiquecyto} → “cytotoxique” 6 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Relevance of compositional translation 7 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Relevance of compositional translation More than 60% of terms in technical and scientific domains are morphologically complex [Namer and Baud, 2007] 7 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Relevance of compositional translation More than 60% of terms in technical and scientific domains are morphologically complex [Namer and Baud, 2007] Outperforms distributional approach for the translation of terms with compositional meaning [Morin and Daille, 2009] 7 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase Elektronen+mikroskop →electron microscope 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase Elektronen+mikroskop →electron microscope ⇒ Restricted to a small set of source-to-target structures 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase Elektronen+mikroskop →electron microscope ⇒ Restricted to a small set of source-to-target structures ⇒ Fertility handled in the specific case of noun compounds 8 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I 9 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I Addressing fertility by allowing translation equivalences from bound morpheme to autonomous lexical item: 9 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I Addressing fertility by allowing translation equivalences from bound morpheme to autonomous lexical item: cyto → cellule ’cell’ 9 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I Addressing fertility by allowing translation equivalences from bound morpheme to autonomous lexical item: cyto → cellule ’cell’ cytotoxic → toxique (pour les) cellules ’toxic to the cells’ 9 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II 10 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II Larger variety of input/output structures: SOURCE TARGET 10 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II Larger variety of input/output structures: SOURCE   prefixed word     neoclassical compound        suffixed word compound      any combination        TARGET 10 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II Larger variety of input/output structures: SOURCE TARGET     prefixed word  prefixed word     neoclassical compound     neoclassical compound              suffixed word suffixed word =⇒ compound  compound          any combination  any combination              phrase 10 / 28
Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 11 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 11 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 2 Translation dictionary look-up 11 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 2 Translation dictionary look-up 3 Recomposition permutations 11 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 2 Translation dictionary look-up 3 Recomposition permutations 4 Selection search occurrences in target texts 11 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 12 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: 12 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens 12 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens match substrings of the source term with: a list of morphemes (prefixes, confixes, suffixes) a list of lexical items 12 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens match substrings of the source term with: a list of morphemes (prefixes, confixes, suffixes) a list of lexical items respect some length constraints on the substrings 12 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens match substrings of the source term with: a list of morphemes (prefixes, confixes, suffixes) a list of lexical items respect some length constraints on the substrings non-cytotoxic → {non, cyto, toxic} 12 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 13 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 Generate all possible concatenations of the minimal components: 13 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 Generate all possible concatenations of the minimal components: { non, cyto, toxic} → {non, cyto, toxic}, {noncyto, toxic}, {non, cytotoxic}, {noncytotoxic} 13 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 Generate all possible concatenations of the minimal components: { non, cyto, toxic} → {non, cyto, toxic}, {noncyto, toxic}, {non, cytotoxic}, {noncytotoxic} ⇒ Increases the chances of matching the components with entries of the dictionaries 13 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up 14 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up Bilingual dictionary for lexical items: toxic → toxique 14 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up Bilingual dictionary for lexical items: toxic → toxique Morpheme translation table for bound morphemes: -cyto- → -cyto-, cellule 14 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up Bilingual dictionary for lexical items: toxic → toxique Morpheme translation table for bound morphemes: -cyto- → -cyto-, cellule {-cyto-, toxic} → {-cyto-, toxique}, {cellule, toxique} 14 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation 15 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation Morphological lexicon toxique → toxicit´ ’toxicity’ e 15 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation Morphological lexicon toxique → toxicit´ ’toxicity’ e Synonyms toxique → v´n´neux ’poisonous’ e e 15 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation Morphological lexicon toxique → toxicit´ ’toxicity’ e Synonyms toxique → v´n´neux ’poisonous’ e e {-cyto-, toxic} → {-cyto-, toxicit´}, e {-cyto-, v´n´neux}, {cellule, toxicit´}, e e e {cellule, v´n´neux} e e 15 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 16 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : 16 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : {-cyto-, toxique} → {-cyto-, toxique}, {toxique, -cyto-} 16 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : {-cyto-, toxique} → {-cyto-, toxique}, {toxique, -cyto-} Recreate target words by generating all possible concatenations of the components : 16 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : {-cyto-, toxique} → {-cyto-, toxique}, {toxique, -cyto-} Recreate target words by generating all possible concatenations of the components : {-cyto-, toxique} → {cytotoxique} {cyto toxique}, 16 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 2 17 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 2 Filter out impossible target terms words e.g.. : “cyto” is a bound morpheme, cannot occur as an autonomous item 17 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 2 Filter out impossible target terms words e.g.. : “cyto” is a bound morpheme, cannot occur as an autonomous item {cyto toxique}, {cytotoxique}→ {cytotoxique} 17 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection 18 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection Match target term with the words of the target corpus 18 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection Match target term with the words of the target corpus Allow at maximum 3 stop words between two words 18 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection Match target term with the words of the target corpus Allow at maximum 3 stop words between two words {toxique cellule} → ‘‘toxique pour les cellules’’ ’toxic to the cells’ 18 / 28
Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 1 http://www.temis.com 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 Comparability [Bo and Gaussier, 2010]: unrelated 0 ⇔ 1 perfectly comparable 1 http://www.temis.com 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 Comparability [Bo and Gaussier, 2010]: unrelated 0 ⇔ 1 perfectly comparable English-French: 0.71 1 http://www.temis.com 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 Comparability [Bo and Gaussier, 2010]: unrelated 0 ⇔ 1 perfectly comparable English-French: 0.71 English-German: 0.45 1 http://www.temis.com 19 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Source terms 20 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Source terms Morphologically constructed word collected from the English texts 20 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Source terms Morphologically constructed word collected from the English texts None of them have a translation in the general language dictionary which is attested in the target texts English to French: 1839 source terms English to German: 1824 source terms 20 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation 21 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) 21 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] 21 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] Morpheme translation table (hand-crafted) 21 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] Morpheme translation table (hand-crafted) Synonyms (Xelda) 21 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] Morpheme translation table (hand-crafted) Synonyms (Xelda) Morphological families [Porter, 1980] 21 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Evaluation measures I Coverage C= σ(STi ) = 1 0 |ST | i=1 σ(STi ) |ST | if |Trans(STi )| ≥ 1 else ⇒ % of source terms with at least 1 translation (regardless of its accuracy) 22 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Evaluation measures II Precision P= |Exact| |Trans| ⇒ % of generated translations which are exact translations Overall quality OQ = C × P ⇒ trade-off between precision and coverage 23 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Experiments 24 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Experiments combination of linguistic resources 24 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Experiments combination of linguistic resources quality of the lexicon with and without the fertile translations 24 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Results: English → French C Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. avg. gain +S +M +D +SMD P -f +f .04 .12 .05 .15 .11 .23 .16 .26 .24 .39 +11 -f +f .81 .57 .69 .50 .20 .28 .70 .60 .31 .33 -8.6 OQ -f +f .03 .07 .03 .08 .02 .06 .11 .16 .07 .13 +4.8 25 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Results: English → German C Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. avg. gain +S +M +D +SMD P -f +f .06 .13 .08 .16 .12 .22 .17 .26 .24 .36 +9.2 -f +f .80 .35 .69 .31 .40 .23 .65 .39 .43 .27 -28.4 OQ -f +f .05 .05 .05 .05 .05 .05 .11 .10 .10 .10 -0.2 26 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results 27 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) 27 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) Morphological types: 27 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) Morphological types: German germanic language: tendency to agglutination oestrogen-independant → Ostrogen-unabh¨ngige a 27 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) Morphological types: German germanic language: tendency to agglutination oestrogen-independant → Ostrogen-unabh¨ngige a French romance language: creates phrases more easily oestrogen-independant → ind´pendant des œstrog`nes e e 27 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis 28 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering 28 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering self-examination → untersuchung selbst ’examination self’ 28 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering self-examination → untersuchung selbst ’examination self’ Wrong or innapropriate translations 28 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering self-examination → untersuchung selbst ’examination self’ Wrong or innapropriate translations in-patient → pas malade ’not ill’ in → “inside” → inside patient in → “inverse” → not a patient 28 / 28
Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work 29 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources 29 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming 29 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming use of a thesaurus 29 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming use of a thesaurus Try translations patterns instead of permutations 29 / 28
Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming use of a thesaurus Try translations patterns instead of permutations Rank translations 29 / 28
Thank you for your attention. B estelle.delpech@univ-nantes.fr beatrice.daille@univ-nantes.fr emmanuel.morin@univ-nantes.fr cl@lingua-et-machina.com
ADDITIONAL SLIDES
Exact translations Non fertiles: pathophysiological → physiopathologique overactive → uberaktiv ¨ Fertiles: cardiotoxicity → toxicit´ cardiaque ’cardiac toxicity’ e mastectomy → ablation der brust ’ablation of the breast’
Morphological variants Non fertiles: dosimetry → dosim´trique ’dosimetric’ e radiosensitivity → strahlenempfindlich ’radiosensitive’ Fertiles: milk-producing → production de lait ’production of milk’ selfexamination → selbst untersuchen ’self examine’
Inexact but semantically related Non fertiles: oncogene → oncog´n`se ’oncogenesis’ e e breakthrough → durchbrechen ’break’ Fertiles: chemoradiotherapy → chemotherapie oder strahlen ’chemotherapy or radiation’ treatable → pouvoir le traiter ’can treat it’
Wrong translations Non fertiles: immunoscore → immunomarquer ’immunostain’ check-in → unkontrollieren ’uncontrolled’ Fertiles: bloodstream → fliessen mehr blut ’more blood flow’ risk-reducing → risque de r´duire ’risk of reducing’ e
R´f´rences I ee Bo, L. and Gaussier, E. (2010). Improving corpus comparability for bilingual lexicon extraction from comparable corpora. In 23`me International Conference on Computational Linguistics, pages 23–27, Beijing, Chine. e Cartoni, B. (2009). Lexical morphology in machine translation: A feasibility study. In Proceedings of the 12th Conference of the European Chapter of the ACL, pages 130–138, Athens, Greece. Harastani, R., Daille, B., and Morin, E. (2012). Neoclassical compound alignments from comparable corpora. In Proceedings of the 13th International Conference on Computational Linguistics and Intelligent Text Processing, volume 2, pages 72–82, New Delhi, India. Hauer, B. and Kondrak, G. (2011). Clustering semantically equivalent words into cognate sets in multilingual lists. In Proceedings of the 5th International Joint Conference on Natural Language Processing, pages 865–873, Chiang Mai, Thailand. Keenan, E. L. and Faltz, L. M. (1985). Boolean semantics for natural language. D. Reidel, Dordrecht, Holland. Morin, E. and Daille, B. (2009). Compositionality and lexical alignment of multi-word terms. In Language Resources and Evaluation (LRE), volume 44 of Multiword expression: hard going or plain sailing, pages 79–95. P. Rayson, S. Piao, S. Sharoff, S. Evert, B. Villada Moir´n, springer netherlands o edition. Namer, F. and Baud, R. (2007). Defining and relating biomedical terms: Towards a cross-language morphosemantics-based system. International Journal of Medical Informatics, 76(2-3):226–33.
R´f´rences II ee Porter, M. F. (1980). An algorithm for suffix stripping. Program, 14(3):130–137. Weller, M., Gojun, A., Heid, U., Daille, B., and Harastani, R. (2011). Simple methods for dealing with term variation and term alignment. In Proceedings of the 9th International Conference on Terminology and Artificial Intelligence, pages 87–93, Paris, France.

More Related Content

PDF
ON THE UTILITY OF A SYLLABLE-LIKE SEGMENTATION FOR LEARNING A TRANSLITERATION...
cscpconf
 
PDF
5. manuel arcedillo & juanjo arevalillo (hermes) translation memories
RIILP
 
PDF
Generic Programming
Ganesh Samarthyam
 
PPTX
Cp 111 lecture 3
HafidhyMasoud
 
PDF
Ceis 3
Alexander Decker
 
PPTX
[Paper Reading] Supervised Learning of Universal Sentence Representations fro...
Hiroki Shimanaka
 
PDF
2. Constantin Orasan (UoW) EXPERT Introduction
RIILP
 
PDF
Nlp presentation
Surya Sg
 
ON THE UTILITY OF A SYLLABLE-LIKE SEGMENTATION FOR LEARNING A TRANSLITERATION...
cscpconf
 
5. manuel arcedillo & juanjo arevalillo (hermes) translation memories
RIILP
 
Generic Programming
Ganesh Samarthyam
 
Cp 111 lecture 3
HafidhyMasoud
 
Ceis 3
Alexander Decker
 
[Paper Reading] Supervised Learning of Universal Sentence Representations fro...
Hiroki Shimanaka
 
2. Constantin Orasan (UoW) EXPERT Introduction
RIILP
 
Nlp presentation
Surya Sg
 

What's hot (20)

PPTX
Nlp (1)
SubramanianMuthusamy3
 
PDF
Nlp research presentation
Surya Sg
 
PDF
Improvement wsd dictionary using annotated corpus and testing it with simplif...
csandit
 
PDF
THE ABILITY OF WORD EMBEDDINGS TO CAPTURE WORD SIMILARITIES
kevig
 
PDF
An investigation of diachronic change in hypotaxis and parataxis in German th...
Mario Bisiada
 
DOCX
Proposal of an Ontology Applied to Technical Debt on PL/SQL Development
Jorge Barreto
 
PPTX
NLP Project Presentation
Aryak Sengupta
 
PDF
IRJET- Short-Text Semantic Similarity using Glove Word Embedding
IRJET Journal
 
PDF
Unsupervised Quality Estimation Model for English to German Translation and I...
Lifeng (Aaron) Han
 
PDF
Latent Topic-semantic Indexing based Automatic Text Summarization
Elaheh Barati
 
PDF
Corpus-based part-of-speech disambiguation of Persian
IDES Editor
 
PPT
Arabic MT Project
Hind Abdulkhaleq
 
PDF
Code Mixing computationally bahut challenging hai
IIIT Hyderabad
 
PDF
ijcai11
Bhaskar Chatterjee
 
PDF
Transformers to Learn Hierarchical Contexts in Multiparty Dialogue
Jinho Choi
 
PDF
Presentation
Pranava Swaroop
 
PDF
RULE BASED TRANSLITERATION SCHEME FOR ENGLISH TO PUNJABI
ijnlc
 
PPTX
Lfg and gpsg
SubramanianMuthusamy3
 
PDF
BERT: Bidirectional Encoder Representations from Transformers
Liangqun Lu
 
PDF
Icpc13.ppt
Ptidej Team
 
Nlp (1)
SubramanianMuthusamy3
 
Nlp research presentation
Surya Sg
 
Improvement wsd dictionary using annotated corpus and testing it with simplif...
csandit
 
THE ABILITY OF WORD EMBEDDINGS TO CAPTURE WORD SIMILARITIES
kevig
 
An investigation of diachronic change in hypotaxis and parataxis in German th...
Mario Bisiada
 
Proposal of an Ontology Applied to Technical Debt on PL/SQL Development
Jorge Barreto
 
NLP Project Presentation
Aryak Sengupta
 
IRJET- Short-Text Semantic Similarity using Glove Word Embedding
IRJET Journal
 
Unsupervised Quality Estimation Model for English to German Translation and I...
Lifeng (Aaron) Han
 
Latent Topic-semantic Indexing based Automatic Text Summarization
Elaheh Barati
 
Corpus-based part-of-speech disambiguation of Persian
IDES Editor
 
Arabic MT Project
Hind Abdulkhaleq
 
Code Mixing computationally bahut challenging hai
IIIT Hyderabad
 
ijcai11
Bhaskar Chatterjee
 
Transformers to Learn Hierarchical Contexts in Multiparty Dialogue
Jinho Choi
 
Presentation
Pranava Swaroop
 
RULE BASED TRANSLITERATION SCHEME FOR ENGLISH TO PUNJABI
ijnlc
 
Lfg and gpsg
SubramanianMuthusamy3
 
BERT: Bidirectional Encoder Representations from Transformers
Liangqun Lu
 
Icpc13.ppt
Ptidej Team
 
Ad

Viewers also liked (17)

PPTX
Cross-lingual ontology lexicalisation, translation and information extraction...
Tobias Wunner
 
PDF
Philippe Langlais - 2017 - Users and Data: The Two Neglected Children of Bili...
Association for Computational Linguistics
 
PDF
Applicative evaluation of bilingual terminologies
Estelle Delpech
 
PDF
Michael Bloodgood - 2017 - Acquisition of Translation Lexicons for Historical...
Association for Computational Linguistics
 
PDF
Chelo Vargas-Sierra
Chelo Vargas
 
PDF
Enriching Transliteration Lexicon Using Automatic Transliteration Extraction
Sarvnaz Karimi
 
PDF
Meng Zhang - 2017 - Adversarial Training for Unsupervised Bilingual Lexicon I...
Association for Computational Linguistics
 
PDF
Dealing with Lexicon Acquired from Comparable Corpora: post-edition and exchange
Estelle Delpech
 
PPTX
A cognitive view of the bilingual lexicon
İrem Tümer
 
PDF
Bilingual terminology mining
Estelle Delpech
 
PDF
Bilingual Terminology Extraction based on Translation Patterns
Alberto Simões
 
PDF
Embedded Human Computation for Knowledge Extraction and Evaluation
webLyzard technology
 
PPTX
Macro economische analyse van brazilië
Jan-Willem Lammens
 
PDF
Parallel text extraction from multimodal comparable corpora
Haithem Afli
 
PDF
Challenges in the linguistic exploitation of specialized republishable web co...
Adrien Barbaresi
 
PDF
Extraction of domain-specific bilingual lexicon from comparable corpora: comp...
Estelle Delpech
 
PPT
Word Formation in English
teflang
 
Cross-lingual ontology lexicalisation, translation and information extraction...
Tobias Wunner
 
Philippe Langlais - 2017 - Users and Data: The Two Neglected Children of Bili...
Association for Computational Linguistics
 
Applicative evaluation of bilingual terminologies
Estelle Delpech
 
Michael Bloodgood - 2017 - Acquisition of Translation Lexicons for Historical...
Association for Computational Linguistics
 
Chelo Vargas-Sierra
Chelo Vargas
 
Enriching Transliteration Lexicon Using Automatic Transliteration Extraction
Sarvnaz Karimi
 
Meng Zhang - 2017 - Adversarial Training for Unsupervised Bilingual Lexicon I...
Association for Computational Linguistics
 
Dealing with Lexicon Acquired from Comparable Corpora: post-edition and exchange
Estelle Delpech
 
A cognitive view of the bilingual lexicon
İrem Tümer
 
Bilingual terminology mining
Estelle Delpech
 
Bilingual Terminology Extraction based on Translation Patterns
Alberto Simões
 
Embedded Human Computation for Knowledge Extraction and Evaluation
webLyzard technology
 
Macro economische analyse van brazilië
Jan-Willem Lammens
 
Parallel text extraction from multimodal comparable corpora
Haithem Afli
 
Challenges in the linguistic exploitation of specialized republishable web co...
Adrien Barbaresi
 
Extraction of domain-specific bilingual lexicon from comparable corpora: comp...
Estelle Delpech
 
Word Formation in English
teflang
 
Ad

Similar to Identification of Fertile Translations in Comparable Corpora: a Morpho-Compositional Approach (20)

PDF
PDFTextProcessing
Joshua Mathias
 
PDF
Two Level Disambiguation Model for Query Translation
IJECEIAES
 
PDF
Unsupervised Software-Specific Morphological Forms Inference from Informal Di...
Chunyang Chen
 
PDF
Neural machine translation of rare words with subword units
Tae Hwan Jung
 
PDF
Parafraseo-Chenggang.pdf
Universidad Nacional de San Martin
 
PPTX
Barreiro-Batista-LR4NLP@Coling2018-presentation
INESC-ID (Spoken Language Systems Laboratory - L2F)
 
PDF
Word2vec on the italian language: first experiments
Vincenzo Lomonaco
 
PDF
Tamil-English Document Translation Using Statistical Machine Translation Appr...
baskaran_md
 
PPTX
More on Indexing Text Operations (1).pptx
Mahsadelavari
 
PDF
Machine Transalation.pdf
Amir Abdalla
 
PPT
Psychological test adaptation
Carlo Magno
 
PPTX
Machine translation from English to Hindi
Rajat Jain
 
PDF
The Effect of Translationese on Statistical Machine Translation
Gennadi Lembersky
 
PDF
Techniques for automatically correcting words in text
unyil96
 
PDF
Lecture: Context-Free Grammars
Marina Santini
 
PPT
A Study Of Statistical Models For Query Translation ：Finding A Good Unit Of T...
iyo
 
PDF
TRANSLATING LEGAL SENTENCE BY SEGMENTATION AND RULE SELECTION
ijnlc
 
PDF
TRANSLATING LEGAL SENTENCE BY SEGMENTATION AND RULE SELECTION
IJNLC Int.Jour on Natural Lang computing
 
PPTX
Academic writing: pointers for G-cube PhD students 23-01.12
Lawrie Hunter
 
PDF
Noun Paraphrasing Based on a Variety of Contexts
Tomoyuki Kajiwara
 
PDFTextProcessing
Joshua Mathias
 
Two Level Disambiguation Model for Query Translation
IJECEIAES
 
Unsupervised Software-Specific Morphological Forms Inference from Informal Di...
Chunyang Chen
 
Neural machine translation of rare words with subword units
Tae Hwan Jung
 
Parafraseo-Chenggang.pdf
Universidad Nacional de San Martin
 
Barreiro-Batista-LR4NLP@Coling2018-presentation
INESC-ID (Spoken Language Systems Laboratory - L2F)
 
Word2vec on the italian language: first experiments
Vincenzo Lomonaco
 
Tamil-English Document Translation Using Statistical Machine Translation Appr...
baskaran_md
 
More on Indexing Text Operations (1).pptx
Mahsadelavari
 
Machine Transalation.pdf
Amir Abdalla
 
Psychological test adaptation
Carlo Magno
 
Machine translation from English to Hindi
Rajat Jain
 
The Effect of Translationese on Statistical Machine Translation
Gennadi Lembersky
 
Techniques for automatically correcting words in text
unyil96
 
Lecture: Context-Free Grammars
Marina Santini
 
A Study Of Statistical Models For Query Translation ：Finding A Good Unit Of T...
iyo
 
TRANSLATING LEGAL SENTENCE BY SEGMENTATION AND RULE SELECTION
ijnlc
 
TRANSLATING LEGAL SENTENCE BY SEGMENTATION AND RULE SELECTION
IJNLC Int.Jour on Natural Lang computing
 
Academic writing: pointers for G-cube PhD students 23-01.12
Lawrie Hunter
 
Noun Paraphrasing Based on a Variety of Contexts
Tomoyuki Kajiwara
 

More from Estelle Delpech (15)

PDF
Génération automatique de texte
Estelle Delpech
 
PDF
Identification de compatibilités entre tages descriptifs de lieux
Estelle Delpech
 
PDF
Découverte du Traitement Automatique des Langues
Estelle Delpech
 
PDF
Invited speaker, ATALA 2014 Ph. D. Thesis award
Estelle Delpech
 
PDF
Corpus comparables et traduction assistée par ordinateur, contributions à la ...
Estelle Delpech
 
PDF
Identification de compatibilites sémantiques entre descripteurs de lieux
Estelle Delpech
 
PDF
Usage du TAL dans des applications industrielles : gestion des contenus multi...
Estelle Delpech
 
PDF
Nomao: data analysis for personalized local search
Estelle Delpech
 
PDF
Nomao: carnet de bonnes adresses (entre amis)
Estelle Delpech
 
PDF
Nomao: local search and recommendation engine
Estelle Delpech
 
PDF
Évaluation applicative des terminologies destinées à la traduction spécialisée
Estelle Delpech
 
PDF
R&D Lingua et Machina
Estelle Delpech
 
PDF
Robust rule-based parsing
Estelle Delpech
 
PDF
Experimenting the TextTiling Algorithm
Estelle Delpech
 
PDF
Text Processing for Procedural Question Answering
Estelle Delpech
 
Génération automatique de texte
Estelle Delpech
 
Identification de compatibilités entre tages descriptifs de lieux
Estelle Delpech
 
Découverte du Traitement Automatique des Langues
Estelle Delpech
 
Invited speaker, ATALA 2014 Ph. D. Thesis award
Estelle Delpech
 
Corpus comparables et traduction assistée par ordinateur, contributions à la ...
Estelle Delpech
 
Identification de compatibilites sémantiques entre descripteurs de lieux
Estelle Delpech
 
Usage du TAL dans des applications industrielles : gestion des contenus multi...
Estelle Delpech
 
Nomao: data analysis for personalized local search
Estelle Delpech
 
Nomao: carnet de bonnes adresses (entre amis)
Estelle Delpech
 
Nomao: local search and recommendation engine
Estelle Delpech
 
Évaluation applicative des terminologies destinées à la traduction spécialisée
Estelle Delpech
 
R&D Lingua et Machina
Estelle Delpech
 
Robust rule-based parsing
Estelle Delpech
 
Experimenting the TextTiling Algorithm
Estelle Delpech
 
Text Processing for Procedural Question Answering
Estelle Delpech
 

Recently uploaded (20)

PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
PPTX
Machine Learning_overview_presentation.pptx
kondavamsidharreddy2
 
PDF
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
Safe Software
 
PDF
MIND Revenue Release Quarter 2 2025 Press Release
MIND CTI
 
PPTX
20250228 LYD VKU AI Blended-Learning.pptx
DatVoNgoc
 
PDF
Reach Out and Touch Someone: Haptics and Empathic Computing
Mark Billinghurst
 
PDF
Unlocking AI with Model Context Protocol (MCP)
Brian McKeiver
 
PPT
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
PDF
Agricultural_Statistics_at_a_Glance_2022_0.pdf
SandeepSingh286037
 
PDF
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
PPTX
Digital-Transformation-Roadmap-for-Companies.pptx
David Malick Dieng
 
PDF
The Rise and Fall of 3GPP – Time for a Sabbatical?
3G4G
 
PPTX
VMware vSphere Foundation How to Sell Presentation-Ver1.4-2-14-2024.pptx
blackmambaettijean
 
PDF
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
PDF
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
SOFTTECHHUB
 
PDF
Diabetes mellitus diagnosis method based random forest with bat algorithm
IAESIJAI
 
PDF
NewMind AI Weekly Chronicles - August'25-Week II
NewMind AI
 
PPTX
Programs and apps: productivity, graphics, security and other tools
4mqw9zch22
 
PDF
Assigned Numbers - 2025 - Bluetooth® Document
Bloombase
 
PDF
Spectral efficient network and resource selection model in 5G networks
IAESIJAI
 
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
Machine Learning_overview_presentation.pptx
kondavamsidharreddy2
 
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
Safe Software
 
MIND Revenue Release Quarter 2 2025 Press Release
MIND CTI
 
20250228 LYD VKU AI Blended-Learning.pptx
DatVoNgoc
 
Reach Out and Touch Someone: Haptics and Empathic Computing
Mark Billinghurst
 
Unlocking AI with Model Context Protocol (MCP)
Brian McKeiver
 
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
Agricultural_Statistics_at_a_Glance_2022_0.pdf
SandeepSingh286037
 
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
Digital-Transformation-Roadmap-for-Companies.pptx
David Malick Dieng
 
The Rise and Fall of 3GPP – Time for a Sabbatical?
3G4G
 
VMware vSphere Foundation How to Sell Presentation-Ver1.4-2-14-2024.pptx
blackmambaettijean
 
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
SOFTTECHHUB
 
Diabetes mellitus diagnosis method based random forest with bat algorithm
IAESIJAI
 
NewMind AI Weekly Chronicles - August'25-Week II
NewMind AI
 
Programs and apps: productivity, graphics, security and other tools
4mqw9zch22
 
Assigned Numbers - 2025 - Bluetooth® Document
Bloombase
 
Spectral efficient network and resource selection model in 5G networks
IAESIJAI
 

Identification of Fertile Translations in Comparable Corpora: a Morpho-Compositional Approach

  • 1. Identification of Fertile Translations in Comparable Corpora a Morpho-Compositional Approach Estelle Delpech1 , B´atrice Daille1 , Emmanuel Morin1 , Claire e Lemaire2,3 1 LINA, 2 GREMUTS, Universit´ de Grenoble Universit´ de Nantes e e 3 Lingua et Machina AMTA’12 10/31/12 San Diego, CA
  • 2. Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
  • 3. Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
  • 4. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context 1 / 28
  • 5. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context Research partly funded by Computer-Aided Translation company 1 / 28
  • 6. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context Research partly funded by Computer-Aided Translation company Goal: generate domain-specific bilingual lexicons when no parallel data is available 1 / 28
  • 7. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Context Research partly funded by Computer-Aided Translation company Goal: generate domain-specific bilingual lexicons when no parallel data is available Available data: general language bilingual dictionary domain-specific comparable corpora 1 / 28
  • 8. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs 2 / 28
  • 9. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: 2 / 28
  • 10. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts 2 / 28
  • 11. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts mixed text types : technical, scientific, lay science... 2 / 28
  • 12. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts mixed text types : technical, scientific, lay science... ⇒ do not expect parallelism in source ↔ target structures 2 / 28
  • 13. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Comparable corpora Definition of comparable corpora Set of texts in languages L1 and L2, which are not translations, but which deal with the same subject matter, so that there is still a possibility to extract translation pairs Some difficulties: language in target texts is not influenced by source texts mixed text types : technical, scientific, lay science... ⇒ do not expect parallelism in source ↔ target structures ⇒ need to deal with variation in translation 2 / 28
  • 14. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation 3 / 28
  • 15. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e 3 / 28
  • 16. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons 3 / 28
  • 17. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance 3 / 28
  • 18. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance ⇒ use of synonyms, thesaurus 3 / 28
  • 19. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance ⇒ use of synonyms, thesaurus Fertility: bi-dimensional → deux dimensions ’two dimensions’ 3 / 28
  • 20. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Variation in translation Morphological variation: anticancer (Noun) → anticanc´reux (Adj) ’anticancerous’ e ⇒ use of morphological derivation rules / lexicons Lexical variation: radiosensitivity → Radiotoleranz ’radiotolerance’ sensitivity ≈ tolerance ⇒ use of synonyms, thesaurus Fertility: bi-dimensional → deux dimensions ’two dimensions’ ⇒ scarcely adressed 3 / 28
  • 21. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility 4 / 28
  • 22. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility Definition target term has more words than the source term 4 / 28
  • 23. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility Definition target term has more words than the source term semantic fertility target term has more morphemes than the source term voie de glace ’route of ice’ → ice climbing route aquarelle (not decomposable) → water color 4 / 28
  • 24. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Original problem Comparable corpora Variation in translation Fertility Definition target term has more words than the source term semantic fertility target term has more morphemes than the source term voie de glace ’route of ice’ → ice climbing route aquarelle (not decomposable) → water color surface fertility target and source terms have the same number of morphemes bi-dimensional → deux dimensions ’two dimensions’ 4 / 28
  • 25. Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
  • 26. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation Principle of compositionality “the meaning of the whole is a function of the meaning of the parts” [Keenan and Faltz, 1985, 24-25] 5 / 28
  • 27. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation Principle of compositionality “the meaning of the whole is a function of the meaning of the parts” [Keenan and Faltz, 1985, 24-25] Definition of compositional translation The translation of the whole is a function of the translation of the parts 5 / 28
  • 28. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 6 / 28
  • 29. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 6 / 28
  • 30. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 2 Translation {cyto, toxic} → {cyto, toxique} 6 / 28
  • 31. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 2 Translation {cyto, toxic} → {cyto, toxique} 3 Recomposition {cyto, toxique} → {cytotoxique, toxiquecyto} 6 / 28
  • 32. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Compositional translation process 1 Decomposition “cytotoxic” → {cyto, toxic} 2 Translation {cyto, toxic} → {cyto, toxique} 3 Recomposition {cyto, toxique} → {cytotoxique, toxiquecyto} 4 Selection {cytotoxique, toxiquecyto} → “cytotoxique” 6 / 28
  • 33. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Relevance of compositional translation 7 / 28
  • 34. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Relevance of compositional translation More than 60% of terms in technical and scientific domains are morphologically complex [Namer and Baud, 2007] 7 / 28
  • 35. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Relevance of compositional translation More than 60% of terms in technical and scientific domains are morphologically complex [Namer and Baud, 2007] Outperforms distributional approach for the translation of terms with compositional meaning [Morin and Daille, 2009] 7 / 28
  • 36. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] 8 / 28
  • 37. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word 8 / 28
  • 38. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e 8 / 28
  • 39. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] 8 / 28
  • 40. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound 8 / 28
  • 41. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e 8 / 28
  • 42. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] 8 / 28
  • 43. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase 8 / 28
  • 44. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase Elektronen+mikroskop →electron microscope 8 / 28
  • 45. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase Elektronen+mikroskop →electron microscope ⇒ Restricted to a small set of source-to-target structures 8 / 28
  • 46. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Related work: single-word terms translation [Cartoni, 2009] Prefixed word → prefixed word ri+organizzare → r´+organiser ’reorganize’ e [Harastani et al., 2012] Neoclassical compound → neoclassical compound Kalori+metrie → calori+m´trie ’calorimetry’ e [Weller et al., 2011] Noun compound → noun phrase Elektronen+mikroskop →electron microscope ⇒ Restricted to a small set of source-to-target structures ⇒ Fertility handled in the specific case of noun compounds 8 / 28
  • 47. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I 9 / 28
  • 48. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I Addressing fertility by allowing translation equivalences from bound morpheme to autonomous lexical item: 9 / 28
  • 49. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I Addressing fertility by allowing translation equivalences from bound morpheme to autonomous lexical item: cyto → cellule ’cell’ 9 / 28
  • 50. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution I Addressing fertility by allowing translation equivalences from bound morpheme to autonomous lexical item: cyto → cellule ’cell’ cytotoxic → toxique (pour les) cellules ’toxic to the cells’ 9 / 28
  • 51. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II 10 / 28
  • 52. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II Larger variety of input/output structures: SOURCE TARGET 10 / 28
  • 53. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II Larger variety of input/output structures: SOURCE   prefixed word     neoclassical compound        suffixed word compound      any combination        TARGET 10 / 28
  • 54. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Underlying principle and advantages Related work Contribution Contribution II Larger variety of input/output structures: SOURCE TARGET     prefixed word  prefixed word     neoclassical compound     neoclassical compound              suffixed word suffixed word =⇒ compound  compound          any combination  any combination              phrase 10 / 28
  • 55. Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
  • 56. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 11 / 28
  • 57. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 11 / 28
  • 58. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 2 Translation dictionary look-up 11 / 28
  • 59. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 2 Translation dictionary look-up 3 Recomposition permutations 11 / 28
  • 60. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Overview 1 Decomposition lexicons + heuristic rules 2 Translation dictionary look-up 3 Recomposition permutations 4 Selection search occurrences in target texts 11 / 28
  • 61. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 12 / 28
  • 62. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: 12 / 28
  • 63. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens 12 / 28
  • 64. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens match substrings of the source term with: a list of morphemes (prefixes, confixes, suffixes) a list of lexical items 12 / 28
  • 65. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens match substrings of the source term with: a list of morphemes (prefixes, confixes, suffixes) a list of lexical items respect some length constraints on the substrings 12 / 28
  • 66. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 1 Split source term into minimal components with heuristic rules: split on hyphens match substrings of the source term with: a list of morphemes (prefixes, confixes, suffixes) a list of lexical items respect some length constraints on the substrings non-cytotoxic → {non, cyto, toxic} 12 / 28
  • 67. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 13 / 28
  • 68. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 Generate all possible concatenations of the minimal components: 13 / 28
  • 69. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 Generate all possible concatenations of the minimal components: { non, cyto, toxic} → {non, cyto, toxic}, {noncyto, toxic}, {non, cytotoxic}, {noncytotoxic} 13 / 28
  • 70. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Decomposition - step 2 Generate all possible concatenations of the minimal components: { non, cyto, toxic} → {non, cyto, toxic}, {noncyto, toxic}, {non, cytotoxic}, {noncytotoxic} ⇒ Increases the chances of matching the components with entries of the dictionaries 13 / 28
  • 71. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up 14 / 28
  • 72. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up Bilingual dictionary for lexical items: toxic → toxique 14 / 28
  • 73. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up Bilingual dictionary for lexical items: toxic → toxique Morpheme translation table for bound morphemes: -cyto- → -cyto-, cellule 14 / 28
  • 74. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation through direct dictionary look-up Bilingual dictionary for lexical items: toxic → toxique Morpheme translation table for bound morphemes: -cyto- → -cyto-, cellule {-cyto-, toxic} → {-cyto-, toxique}, {cellule, toxique} 14 / 28
  • 75. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation 15 / 28
  • 76. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation Morphological lexicon toxique → toxicit´ ’toxicity’ e 15 / 28
  • 77. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation Morphological lexicon toxique → toxicit´ ’toxicity’ e Synonyms toxique → v´n´neux ’poisonous’ e e 15 / 28
  • 78. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Translation with variation Morphological lexicon toxique → toxicit´ ’toxicity’ e Synonyms toxique → v´n´neux ’poisonous’ e e {-cyto-, toxic} → {-cyto-, toxicit´}, e {-cyto-, v´n´neux}, {cellule, toxicit´}, e e e {cellule, v´n´neux} e e 15 / 28
  • 79. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 16 / 28
  • 80. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : 16 / 28
  • 81. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : {-cyto-, toxique} → {-cyto-, toxique}, {toxique, -cyto-} 16 / 28
  • 82. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : {-cyto-, toxique} → {-cyto-, toxique}, {toxique, -cyto-} Recreate target words by generating all possible concatenations of the components : 16 / 28
  • 83. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 1 Permutate the target components : {-cyto-, toxique} → {-cyto-, toxique}, {toxique, -cyto-} Recreate target words by generating all possible concatenations of the components : {-cyto-, toxique} → {cytotoxique} {cyto toxique}, 16 / 28
  • 84. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 2 17 / 28
  • 85. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 2 Filter out impossible target terms words e.g.. : “cyto” is a bound morpheme, cannot occur as an autonomous item 17 / 28
  • 86. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Recomposition - step 2 Filter out impossible target terms words e.g.. : “cyto” is a bound morpheme, cannot occur as an autonomous item {cyto toxique}, {cytotoxique}→ {cytotoxique} 17 / 28
  • 87. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection 18 / 28
  • 88. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection Match target term with the words of the target corpus 18 / 28
  • 89. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection Match target term with the words of the target corpus Allow at maximum 3 stop words between two words 18 / 28
  • 90. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Decomposition Translation Recomposition Selection Selection Match target term with the words of the target corpus Allow at maximum 3 stop words between two words {toxique cellule} → ‘‘toxique pour les cellules’’ ’toxic to the cells’ 18 / 28
  • 91. Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
  • 92. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora 19 / 28
  • 93. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German 19 / 28
  • 94. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer 19 / 28
  • 95. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 19 / 28
  • 96. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science 19 / 28
  • 97. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 1 http://www.temis.com 19 / 28
  • 98. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 Comparability [Bo and Gaussier, 2010]: unrelated 0 ⇔ 1 perfectly comparable 1 http://www.temis.com 19 / 28
  • 99. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 Comparability [Bo and Gaussier, 2010]: unrelated 0 ⇔ 1 perfectly comparable English-French: 0.71 1 http://www.temis.com 19 / 28
  • 100. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Corpora English, French, German breast cancer approx. 400k words per language 1 2 scientic papers + 1 2 lay science pos-tagged with software Xelda1 Comparability [Bo and Gaussier, 2010]: unrelated 0 ⇔ 1 perfectly comparable English-French: 0.71 English-German: 0.45 1 http://www.temis.com 19 / 28
  • 101. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Source terms 20 / 28
  • 102. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Source terms Morphologically constructed word collected from the English texts 20 / 28
  • 103. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Source terms Morphologically constructed word collected from the English texts None of them have a translation in the general language dictionary which is attested in the target texts English to French: 1839 source terms English to German: 1824 source terms 20 / 28
  • 104. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation 21 / 28
  • 105. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) 21 / 28
  • 106. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] 21 / 28
  • 107. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] Morpheme translation table (hand-crafted) 21 / 28
  • 108. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] Morpheme translation table (hand-crafted) Synonyms (Xelda) 21 / 28
  • 109. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Resources for translation General language dictionary (Xelda) Domain-specific dictionary : cognates extracted from corpus [Hauer and Kondrak, 2011] Morpheme translation table (hand-crafted) Synonyms (Xelda) Morphological families [Porter, 1980] 21 / 28
  • 110. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Evaluation measures I Coverage C= σ(STi ) = 1 0 |ST | i=1 σ(STi ) |ST | if |Trans(STi )| ≥ 1 else ⇒ % of source terms with at least 1 translation (regardless of its accuracy) 22 / 28
  • 111. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Evaluation measures II Precision P= |Exact| |Trans| ⇒ % of generated translations which are exact translations Overall quality OQ = C × P ⇒ trade-off between precision and coverage 23 / 28
  • 112. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Experiments 24 / 28
  • 113. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Experiments combination of linguistic resources 24 / 28
  • 114. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Experiments combination of linguistic resources quality of the lexicon with and without the fertile translations 24 / 28
  • 115. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Results: English → French C Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. avg. gain +S +M +D +SMD P -f +f .04 .12 .05 .15 .11 .23 .16 .26 .24 .39 +11 -f +f .81 .57 .69 .50 .20 .28 .70 .60 .31 .33 -8.6 OQ -f +f .03 .07 .03 .08 .02 .06 .11 .16 .07 .13 +4.8 25 / 28
  • 116. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Results: English → German C Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. Gen.+Morph. avg. gain +S +M +D +SMD P -f +f .06 .13 .08 .16 .12 .22 .17 .26 .24 .36 +9.2 -f +f .80 .35 .69 .31 .40 .23 .65 .39 .43 .27 -28.4 OQ -f +f .05 .05 .05 .05 .05 .05 .11 .10 .10 .10 -0.2 26 / 28
  • 117. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results 27 / 28
  • 118. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) 27 / 28
  • 119. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) Morphological types: 27 / 28
  • 120. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) Morphological types: German germanic language: tendency to agglutination oestrogen-independant → Ostrogen-unabh¨ngige a 27 / 28
  • 121. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Discussion: English-French vs. English-German results English-German corpus is much less comparable (0.45 vs. 0.71) Morphological types: German germanic language: tendency to agglutination oestrogen-independant → Ostrogen-unabh¨ngige a French romance language: creates phrases more easily oestrogen-independant → ind´pendant des œstrog`nes e e 27 / 28
  • 122. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis 28 / 28
  • 123. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering 28 / 28
  • 124. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering self-examination → untersuchung selbst ’examination self’ 28 / 28
  • 125. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering self-examination → untersuchung selbst ’examination self’ Wrong or innapropriate translations 28 / 28
  • 126. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Data Evaluation measures Results Error analysis Problems in word reordering self-examination → untersuchung selbst ’examination self’ Wrong or innapropriate translations in-patient → pas malade ’not ill’ in → “inside” → inside patient in → “inverse” → not a patient 28 / 28
  • 127. Outline 1 Context and original problem 2 Compositional translation framework 3 Detailed translation method 4 Experiments and results 5 Future work
  • 128. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work 29 / 28
  • 129. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources 29 / 28
  • 130. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming 29 / 28
  • 131. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming use of a thesaurus 29 / 28
  • 132. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming use of a thesaurus Try translations patterns instead of permutations 29 / 28
  • 133. Context and original problem Compositional translation framework Detailed translation method Experiments and results Future work Future work Improve quality of linguistic resources morphological derivation rules instead of stemming use of a thesaurus Try translations patterns instead of permutations Rank translations 29 / 28
  • 134. Thank you for your attention. B estelle.delpech@univ-nantes.fr beatrice.daille@univ-nantes.fr emmanuel.morin@univ-nantes.fr cl@lingua-et-machina.com
  • 136. Exact translations Non fertiles: pathophysiological → physiopathologique overactive → uberaktiv ¨ Fertiles: cardiotoxicity → toxicit´ cardiaque ’cardiac toxicity’ e mastectomy → ablation der brust ’ablation of the breast’
  • 137. Morphological variants Non fertiles: dosimetry → dosim´trique ’dosimetric’ e radiosensitivity → strahlenempfindlich ’radiosensitive’ Fertiles: milk-producing → production de lait ’production of milk’ selfexamination → selbst untersuchen ’self examine’
  • 138. Inexact but semantically related Non fertiles: oncogene → oncog´n`se ’oncogenesis’ e e breakthrough → durchbrechen ’break’ Fertiles: chemoradiotherapy → chemotherapie oder strahlen ’chemotherapy or radiation’ treatable → pouvoir le traiter ’can treat it’
  • 139. Wrong translations Non fertiles: immunoscore → immunomarquer ’immunostain’ check-in → unkontrollieren ’uncontrolled’ Fertiles: bloodstream → fliessen mehr blut ’more blood flow’ risk-reducing → risque de r´duire ’risk of reducing’ e
  • 140. R´f´rences I ee Bo, L. and Gaussier, E. (2010). Improving corpus comparability for bilingual lexicon extraction from comparable corpora. In 23`me International Conference on Computational Linguistics, pages 23–27, Beijing, Chine. e Cartoni, B. (2009). Lexical morphology in machine translation: A feasibility study. In Proceedings of the 12th Conference of the European Chapter of the ACL, pages 130–138, Athens, Greece. Harastani, R., Daille, B., and Morin, E. (2012). Neoclassical compound alignments from comparable corpora. In Proceedings of the 13th International Conference on Computational Linguistics and Intelligent Text Processing, volume 2, pages 72–82, New Delhi, India. Hauer, B. and Kondrak, G. (2011). Clustering semantically equivalent words into cognate sets in multilingual lists. In Proceedings of the 5th International Joint Conference on Natural Language Processing, pages 865–873, Chiang Mai, Thailand. Keenan, E. L. and Faltz, L. M. (1985). Boolean semantics for natural language. D. Reidel, Dordrecht, Holland. Morin, E. and Daille, B. (2009). Compositionality and lexical alignment of multi-word terms. In Language Resources and Evaluation (LRE), volume 44 of Multiword expression: hard going or plain sailing, pages 79–95. P. Rayson, S. Piao, S. Sharoff, S. Evert, B. Villada Moir´n, springer netherlands o edition. Namer, F. and Baud, R. (2007). Defining and relating biomedical terms: Towards a cross-language morphosemantics-based system. International Journal of Medical Informatics, 76(2-3):226–33.
  • 141. R´f´rences II ee Porter, M. F. (1980). An algorithm for suffix stripping. Program, 14(3):130–137. Weller, M., Gojun, A., Heid, U., Daille, B., and Harastani, R. (2011). Simple methods for dealing with term variation and term alignment. In Proceedings of the 9th International Conference on Terminology and Artificial Intelligence, pages 87–93, Paris, France.