Bilingual Corpus Research Articles

Bayesian Word Learning in Multiple Language Environments.

Infant language learners are faced with the difficult inductive problem of determining how new words map to novel or known objects in their environment. Bayesian inference models have been successful at using the sparse information available in natural child-directed speech to build candidate lexicons and infer speakers' referential intentions. We begin by asking how a Bayesian model optimized for monolingual input (the Intentional Model; Frank etal., 2009) generalizes to new monolingual or bilingual corpora and find that, especially in the case of the bilingual input, the model shows a significant decrease in performance. In the next experiment, we propose the ME Model, a modified Bayesian model, which approximates infants' mutual exclusivity bias to support the differential demands of monolingual and bilingual learning situations. The extended model is assessed using the same corpora of real child-directed speech, showing that its performance is more robust against varying input and less dependent than the Intentional Model on optimization of its parsimony parameter. We argue that both monolingual and bilingual demands on word learning are important considerations for a computational model, as they can yield significantly different results than when only one such context is considered.

Does empirical data from bilingual and native Spanish corpora meet linguistic theory? The role of discourse context in variation of subject expression

Abstract The goal of this study is to shed light on how empirical data on the discourse constraints of null and overt third person subject pronouns in L1 and bilingual Spanish meet linguistic theory. A (semi)spontaneous production task was administered to 34 Moroccan Arabic (MA)/Spanish early sequential bilinguals and 30 L1 Spanish controls. All 3rd person subject positions were coded: (1) morphosyntactic form (null pronoun vs. overt pronoun); (2) discourse function ([-Topic Shift] vs. [+Topic Shift]); (3) sentence relation (intrasentential vs. intersentential); (4) clause order within intrasentential contexts (main-subordinate vs. subordinate-main); and (5) access to the antecedent (clear vs. ambiguous antecedent). The results reveal general patterns of use in both L1 and bilingual Spanish: null pronouns express topic maintenance both in inter- and intrasentential contexts (both clause orders) and overt pronouns, especially in intersentential contexts, are generally used for topic change. However, additional analyses provide evidence that null pronouns in L1, but not in bilingual Spanish, are often used in change of reference contexts where the antecedent is not ambiguous. This reveals patterns that have gone unreported by most previous descriptive and theoretical studies. Finally, a higher use of ambiguous null pronouns is attested among bilingual speakers, which suggests a lower control of the mechanisms by which reference is established in discourse and supports, to some extent, the predictions derived from the Interface Hypothesis.

Linguistic-Relationships-Based Approach for Improving Word Alignment

The unsupervised word alignments (such as GIZA++) are widely used in the phrase-based statistical machine translation. The quality of the model is proportional to the size and the quality of the bilingual corpus. However, for low-resource language pairs such as Chinese and Vietnamese, a result of unsupervised word alignment sometimes is of low quality due to the sparse data. In addition, this model does not take advantage of the linguistic relationships to improve performance of word alignment. Chinese and Vietnamese have the same language type and have close linguistic relationships. In this article, we integrate the characteristics of linguistic relationships into the word alignment model to enhance the quality of Chinese-Vietnamese word alignment. These linguistic relationships are Sino-Vietnamese and content word. The experimental results showed that our method improved the performance of word alignment as well as the quality of machine translation.

Reframing translated news for target readers: a narrative account of news translation in Snowden’s discourses

ABSTRACTThis article studies some of the framing strategies mentioned in extant literature in the Chinese and English versions of news reports on the Edward Snowden case. It analyses a bilingual corpus of 47 news texts in each language. Both the English and Chinese data were collected from Chinese state-owned news agencies. The research results show that even though the news agency that produced the translated news shares the same ideology as the news organization that provided the Chinese news, the translated narratives of the incident were often reframed. We argue that the possible factors that affect the reframing of the narrative include the target readership, the political situation and the political position of the news agency. We also argue that applying a narrative account may help to better explore stance variation in sensitive news translation.

Synchrony issues in comics. Language transfer and gender-specific characterisation in English translations of Greek Aristophanic comics

ABSTRACTIn the last few years, two veritably burgeoning areas, comics studies and translation studies have asserted their autonomy by addressing specificities of form and context of production/reception. Acknowledging similarities between these two fields, and highlighting the role of translation as a conduit of cultural flows and representations, this article explores linguistic transfer and male/female characterisation in the English translations of Assembly of Women and Ladies’ Day. The two comics are adaptations of Aristophanic playtexts and their translations were launched as part of the general educational mission of a Greek publishing house, Metaichmio. Originals and translations are compared with the help of categories of synchrony, a concept traditionally used in audio-visual translation and adapted here to indicate alignment between text and visuals in translation: kinetic synchrony (movement and gestures), content synchrony (contextual equivalence), isochrony (text volume) and character synchrony (performative preferences for individual characters). Despite a general emphasis on space constraints in the literature, a bilingual comics corpus compiled here shows patterns of creative rewriting affecting characterisation.

Translating Low-Resource Languages by Vocabulary Adaptation from Close Counterparts

Some natural languages belong to the same family or share similar syntactic and/or semantic regularities. This property persuades researchers to share computational models across languages and benefit from high-quality models to boost existing low-performance counterparts. In this article, we follow a similar idea, whereby we develop statistical and neural machine translation (MT) engines that are trained on one language pair but are used to translate another language. First we train a reliable model for a high-resource language, and then we exploit cross-lingual similarities and adapt the model to work for a close language with almost zero resources. We chose Turkish (Tr) and Azeri or Azerbaijani (Az) as the proposed pair in our experiments. Azeri suffers from lack of resources as there is almost no bilingual corpus for this language. Via our techniques, we are able to train an engine for the Az → English (En) direction, which is able to outperform all other existing models.

Translation Quality Estimation Using Only Bilingual Corpora

In computer-aided translation scenarios, quality estimation of machine translation hypotheses plays a critical role. Existing methods for word-level translation quality estimation (TQE) rely on the availability of manually annotated TQE training data obtained via direct annotation or postediting. However, due to the cost of human labor, such data are either limited in size or is only available for few tasks in practice. To avoid the reliance on such annotated TQE data, this paper proposes an approach to train word-level TQE models using bilingual corpora, which are typically used in machine translation training and is relatively easier to access. We formalize the training of our proposed method under the framework of maximum marginal likelihood estimation. To avoid degenerated solutions, we propose a novel regularized training objective whose optimization is achieved by an efficient approximation. Extensive experiments on both written and spoken language datasets empirically show that our approach yields comparable performance to the standard training on annotated data.

Acquiring Chinese paraphrases based on random walk of $N$ steps

The conventional “pivot” approach of acquiring paraphrases from bilingual corpus has certain limitations where only candidate paraphrases within two steps are considered. In this paper, we propose a graph-based model of acquiring paraphrases from a phrase translation table. First, we describe a graph-based model representing Chinese-English phrase translation relations, a random walk algorithm based on $N$ number of steps and a confidence metric for the obtained paraphrases. Furthermore, with the aim of finding more potential for Chinese paraphrases, we augment the model so that it is able to integrate other language pairs, such as English-Japanese phrase translation relations. We performed experiments on NTCIR Chinese-English and English-Japanese bilingual corpus and compared the results to those of conventional methods. The experimental results show that the proposed approach acquires more paraphrases. In addition, the performance was improved further after the English-Japanese phrase translations were added to the graph-based model.

La expresión lingüística de la valoración en textos jurisprudenciales: Estudio contrastivo francés-español

This paper focuses on the linguistic mechanisms which reflect the speaker’s subjective evaluation of textual contents, as well as on how different enunciative voices are intertwined in texts. In order to examine these evaluative mechanisms, Appraisal Theory (Halliday & Mathiessen, 2004; Martin & White, 2005) is applied to a bilingual corpus on liability in health care. The corpus is made up by twenty judicial decisions issued by both the first civil chamber of the French Tribunal de Cassation and the first civil chamber of the Spanish Tribunal Supremo during the first decade of the 21st century. The linguistic resources analysed in the study are those used by the speaker to convey subjective evaluation about the patient’s initial condition, the medical intervention and the results of medical actions. The contrastive analysis has shed light on similarities and differences between French and Spanish related to evaluative linguistic mechanisms and core values that define healthcare responsibility in each culture.

Neural Networks Classifier for Data Selection in Statistical Machine Translation

AbstractCorpora are precious resources, as they allow for a proper estimation of statistical machine translation models. Data selection is a variant of the domain adaptation field, aimed to extract those sentences from an out-of-domain corpus that are the most useful to translate a different target domain. We address the data selection problem in statistical machine translation as a classification task. We present a new method, based on neural networks, able to deal with monolingual and bilingual corpora. Empirical results show that our data selection method provides slightly better translation quality, compared to a state-of-the-art method (cross-entropy), requiring substantially less data. Moreover, the results obtained are coherent across different language pairs, demonstrating the robustness of our proposal.

Cross-lingual neighborhood effects in generalized lexical decision and natural reading.

The present study assessed intra- and cross-lingual neighborhood effects, using both a generalized lexical decision task and an analysis of a large-scale bilingual eye-tracking corpus (Cop, Dirix, Drieghe, & Duyck, 2016). Using new neighborhood density and frequency measures, the general lexical decision task yielded an inhibitory cross-lingual neighborhood density effect on reading times of second language words, replicating van Heuven, Dijkstra, and Grainger (1998). Reaction times for native language words were not influenced by neighborhood density or frequency but error rates showed cross-lingual neighborhood effects depending on target word frequency. The large-scale eye movement corpus confirmed effects of cross-lingual neighborhood on natural reading, even though participants were reading a novel in a unilingual context. Especially second language reading and to a lesser extent native language reading were influenced by lexical candidates from the nontarget language, although these effects in natural reading were largely facilitatory. These results offer strong and direct support for bilingual word recognition models that assume language-independent lexical access. (PsycINFO Database Record

A SYNTACTIC ANALYSIS OF THE ENGLISH DISCOURSE MARKER ONLY AND ITS VIETNAMESE TRANSLATIONAL EQUIVALENTS

This descriptive and qualitative study, part of a big project, endeavored to explore the syntactic features of the English discourse marker only and its Vietnamese translational equivalents. It involved the development of a bilingual corpus, which consisted of 168 English and 168 Vietnamese instances of sentences containing only. The interpretive contrastive analysis was employed. The findings reveal that only in English often comes in three positions, namely, initial, medial, final, whereas their Vietnamese equivalent usually appears in initial and medial positions. Additionally, there are more similarities than differences in terms of syntactic functions of only used in English and Vietnamese.

COMPREHENSIVE APPROACH FOR BILINGUAL MACHINE TRANSLATION

Machine Translation has been a major focus of the NLP group since 1999, the principal focus of the Natural Language Processing group is to build a machine translation system that automatically learns translation mappings from bilingual corpora. This paper explores a novel approach for phrase based machine translation from English to Kannada and Kannada to English. The source text is analyzed then simple sentences are translated using the rules and the complex sentences are split into simple sentences later translation is performed.

The Feasibility of Content and System Morpheme Hierarchy in the Analysis of Tamazight Bilingual Corpora: The Case of Kabyle and Mzabi Bilingual Speech in Oran

This study examines the empirical validity of the hierarchy of system and content morphemes on Tamazight bilingual corpora. This dichotomy is one of the underlying principles of the Matrix Language Frame model and the 4-M model as they have been advocated by Myers- Scotton in 1997, 2002 and 2016. These socio-psychologically based syntactic models have been used by contact linguists as viable alternatives in the investigation and interpretation of the morphosyntactic processes underlying bilingual corpora. The present paper investigates Kabyle and Mzabi bilingual data; it focuses on these two Berber or Tamazight varieties as they are in contact with Algerian Arabic, Standard Arabic and French in an Algerian context, namely Oran city. The study of the communities under light here reveals that the languages composing the verbal repertoire of these Tamazight minorities living in Oran overlap at different linguistic levels including the morpho-syntactic one.

Leveraging bilingually-constrained synthetic data via multi-task neural networks for implicit discourse relation recognition

Recognizing implicit discourse relations is an important but challenging task in discourse understanding. To alleviate the shortage of labeled data, previous work automatically generates synthetic implicit data (SynData) as additional training data, by removing connectives from explicit discourse instances. Although SynData has been proven useful for implicit discourse relation recognition, it also has the meaning shift problem and the domain problem. In this paper, we first propose to use bilingually-constrained synthetic implicit data (BiSynData) to enrich the training data, which can alleviate the drawbacks of SynData. Our BiSynData is constructed from a bilingual sentence-aligned corpus according to the implicit/explicit mismatch between different languages. Then we design a multi-task neural network model to incorporate our BiSynData to benefit implicit discourse relation recognition. Experimental results on both the English PDTB and Chinese CDTB data sets show that our proposed method achieves significant improvements over baselines using SynData.

Building Laos Dependency Treebank by Means of Chinese-Laos Bilingual Corpus of Word Alignment

Building Laos Dependency Treebank by Means of Chinese-Laos Bilingual Corpus of Word Alignment

Towards a new possibilistic query translation tool for cross-language information retrieval

Approaches of query translation in Cross-Language Information Retrieval (CLIR) have frequently used dictionaries which suffer from translation ambiguity. Besides, a word-by-word query translation is not sufficient. In this paper, we propose, evaluate and compare a new possibilistic approach for query translation in order to improve the previous dictionary-based ones. This approach uses a probability-to-possibility transformation as a mean to introduce further tolerance in query translation process. Firstly, we identify noun phrases (NPs) in the source query and translate them as units using translation patterns and a language model. Secondly, source query terms which are not included in any selected NPs are translated word-by-word using our new possibilistic approach of single word translation. Indeed, we take into account all query words and their translations when we choose the suitable translation of a given word. We start from the idea that the correct suitable translations of query terms have a tendency to co-occur in the target language documents unlike unsuitable ones. Finally, to increase the coverage of the bilingual dictionary, additional words and their translations are automatically generated from a parallel bilingual corpus. We tested our approach using the French-English parallel text corpus Europarl and the CLEF-2003 French-English CLIR test collection. The reported experiments showed the performance of the probability-to-possibility transformation-based approach compared to the probabilistic one and to some state-of-the-art CLIR tools.

Translation and Popularization: Medical Research in the Communicative Continuum

Far from being restricted to exchanges between experts, specialised knowledge is mediated to audiences with different levels of specialization, from scientific reviews to newspaper articles. This diversity constitutes an often-overlooked challenge for translators. As a matter of fact, while documentation and terminology are always crucial, translation decisions are based on communicative parameters as well as cognitive and linguistic criteria. Although it is self-evident that linguistic choices are determined by the proficiency level of the readership, few authors have attempted to specify what those choices are and how the correlation operates, most notably in popularization discourse, and none of them has considered potential differences between languages and cultural settings. The focus of the paper is a bilingual (French and Spanish) corpus study carried out on newspaper articles dealing with stem cell research and cloning published in four different geographic regions (France, Quebec, Spain, Argentina). An original methodology was implemented for data collection and analysis. The number and nature of expressions used to convey each concept were then analyzed. Discursive strategies widely assumed to be a hallmark of popularization, like definitions and explanations, were also taken into account. Indices of metaphorical conceptualization and the underlying modes of conceptualization were identified. This study provides concrete data to a debate that remains largely theoretical, and supports the conception of specialized communication as a continuum. The results go against well-established ideas about popularized texts, specially regarding the trademark status of “didactic features.” It seems imperative to acknowledge the heterogeneity of popularization and to consider the role of textual genre constraints in the way specialized knowledge is introduced. Furthermore, the data obtained seems to substantiate the recent questioning of the canonical view of popularization as a mere translation.

POS-tagging a bilingual parallel corpus: methods and challenges

This paper reviews the author’s experiences of tokenizing and POS tagging a bilingual parallel corpus, the PaGeS Corpus, consisting mostly of German and Spanish fictional texts. This is part of an ongoing process of annotating the corpus for part-of-speech information. This study discusses the specific problems encountered so far. On the one hand, tagging performance degrades significantly when applied to fictional data and, on the other, pre-existing annotation schemes are all language specific. To further improve accuracy during post-editing, the author has developed a common tagset and identified major error patterns.

Chinese temporal relation resolution based on Chinese-English parallel corpus

This paper studies the annotation of temporal objects such as EVENT-type and TIMEX3-type elements in Chinese text based on Chinese-English bilingual parallel corpus. We propose and implement a temporal relation resolution (TRR) engine for Chinese text. We use temporal recognition and normalisation tools to annotate the English text, and then map the annotations of English text to the Chinese text by word alignment. The constructed Chinese TimeBank could be used for TRR research based on TimeML in Chinese text.