Bilingual Lexicon Induction From Comparable and Parallel Data: A Comparative Analysis

We present InstaMap, an instance-based method for learning projection-based cross-lingual word embeddings. Unlike prior work, it deviates from learning a single global linear projection. InstaMap is a non-parametric model that learns a non-linear projection by iteratively: (1) finding a globally optimal rotation of the source embedding space relying on the Kabsch algorithm, and then (2) moving each point along an instance-specific translation vector estimated from the translation vectors of the point’s nearest neighbours in the training dictionary. We report performance gains with InstaMap over four representative state-of-the-art projection-based models on bilingual lexicon induction across a set of 28 diverse language pairs. We note prominent improvements, especially for more distant language pairs (i.e., languages with non-isomorphic monolingual spaces).

Neural machine translation (NMT) systems are usually trained on clean parallel data. They can perform very well for translating clean in-domain texts. However, as demonstrated by previous work, the translation quality significantly worsens when translating noisy texts, such as user-generated texts (UGT) from online social media. Given the lack of parallel data of UGT that can be used to train or adapt NMT systems, we synthesize parallel data of UGT, exploiting monolingual data of UGT through cross-lingual language model pre-training and zero-shot NMT systems. This paper presents two different but complementary approaches: one alters given clean parallel data into UGT-like parallel data whereas the other generates translations from monolingual data of UGT. On the MTNT translation tasks, we show that our synthesized parallel data can lead to better NMT systems for UGT while making them more robust in translating texts from various domains and styles.

Among the six challenges of neural machine translation (NMT) coined by (Koehn and Knowles, 2017), rare-word problem is considered the most severe one, especially in translation of low-resource languages. In this paper, we propose three solutions to address the rare words in neural machine translation systems. First, we enhance source context to predict the target words by connecting directly the source embeddings to the output of the attention component in NMT. Second, we propose an algorithm to learn morphology of unknown words for English in supervised way in order to minimize the adverse effect of rare-word problem. Finally, we exploit synonymous relation from the WordNet to overcome out-of-vocabulary (OOV) problem of NMT. We evaluate our approaches on two low-resource language pairs: English-Vietnamese and Japanese-Vietnamese. In our experiments, we have achieved significant improvements of up to roughly +1.0 BLEU points in both language pairs.

This paper describes the sourcing, processing, and application of parallel text data for Icelandic and Polish for the purpose of bilingual lexicon induction (BLI), demonstrating how a parallel corpus can be compiled for a low-to-medium resource language pair that has no available parallel data, by pivoting through a common language. We show the usefulness of the corpus by training and evaluating a machine translation (MT) model on the data. Iceland's linguistic landscape is evolving, with an increasing need for multilingual support due to the growing immigrant population. Polish, in particular, stands out as the language of the largest single minority in Iceland, underscoring the importance of this project.

Most neural machine translation (NMT) systems rely on parallel data, comprising text in the source language and its corresponding translation in the target language. While it’s acknowledged that context enhances NMT models, this work proposes a novel approach by incorporating external context, specifically explanations of source text meanings, akin to how human translators leverage context for comprehension. The suggested methodology innovatively addresses the challenge of incorporating lengthy contextual information into NMT systems. By employing state-of-the-art transformer-based models, external context is integrated, thereby enriching the translation process. A key aspect of the approach lies in the utilization of diverse text summarization techniques, strategically employed to efficiently distill extensive contextual details into the NMT framework. This novel solution not only overcomes the obstacle posed by lengthy context but also enhances the translation quality, marking a advancement in the field of NMT. Furthermore, the data-centric approach ensures robustness and effectiveness, yielding improvements in translation quality, as evidenced by a considerable boost in BLEU score points ranging from 0.46 to 1.87 over baseline models. Additionally, we make our dataset publicly available, facilitating further research in this domain.

Recently, the emergence of the digital language division and the availability of cross-lingual benchmarks make researches of cross-lingual texts more popular. However, the performance of existing methods based on mapping relation are not good enough, because sometimes the structures of language spaces are not isomorphic. Besides, polysemy makes the extraction of interaction features hard. For cross-lingual word embedding, a model named Cross-lingual Word Embedding Space Based on Pseudo Corpus (CWE-PC) is proposed to obtain cross-lingual and multilingual word embedding. For cross-lingual sentence pair interaction feature capture, a Cross-language Feature Capture Based on Similarity Matrix (CFC-SM) model is built to extract cross-lingual interaction features. ELMo pretrained model and multiple layer convolution are used to alleviate polysemy and extract interaction features. These models are evaluated on multiple language pairs and results show that they outperform the state-of-the-art cross-lingual word embedding methods.

Recent research on cross-lingual word embeddings has been dominated by unsupervised mapping approaches that align monolingual embeddings. Such methods critically rely on those embeddings having a similar structure, but it was recently shown that the separate training in different languages causes departures from this assumption. In this paper, we propose an alternative approach that does not have this limitation, while requiring a weak seed dictionary (e.g., a list of identical words) as the only form of supervision. Rather than aligning two fixed embedding spaces, our method works by fixing the target language embeddings, and learning a new set of embeddings for the source language that are aligned with them. To that end, we use an extension of skip-gram that leverages translated context words as anchor points, and incorporates self-learning and iterative restarts to reduce the dependency on the initial dictionary. Our approach outperforms conventional mapping methods on bilingual lexicon induction, and obtains competitive results in the downstream XNLI task.

We participated in the WMT 2018 shared news translation task in three language pairs: English-Estonian, English-Finnish, and English-Czech. Our main focus was the low-resource language pair of Estonian and English for which we utilized Finnish parallel data in a simple method. We first train a “parent model” for the high-resource language pair followed by adaptation on the related low-resource language pair. This approach brings a substantial performance boost over the baseline system trained only on Estonian-English parallel data. Our systems are based on the Transformer architecture. For the English to Czech translation, we have evaluated our last year models of hybrid phrase-based approach and neural machine translation mainly for comparison purposes.

Machine translation aims to minimize the language barrier between people of different linguistic backgrounds. Machine translation is an automatic translation technique between pairs of different languages. Machine translation using neural network or neural machine translation came into the picture due to several limitation associated with its predecessors which are rule based and statistical based models. For large data sets with a rich range of vocabulary, the neural network machine translation system provides fair translation accuracy. We have observed that there remain very few machine translators dedicated to Indian languages, especially those spoken in the North-East area. So our paper mainly focuses on implementing a neural machine translation system for the English-Assamese language. In this paper we used five different neural machine translation models for English-Assamese language pair using LSTMs and GRUs, along with an attention layer. We have made a comparison analysis based on the performance results of these models. We used BLEU Score to calculate the accuracy of these five models, thus achieving a BLEU score of 34.168%, which was the highest among the five models.

In this paper, we deal with two problems in Japanese-English machine translation of news articles. The first problem is the quality of parallel corpora. Neural machine translation (NMT) systems suffer degraded performance when trained with noisy data. Because there is no clean Japanese-English parallel data for news articles, we build a novel parallel news corpus consisting of Japanese news articles translated into English in a content-equivalent manner. This is the first content-equivalent Japanese-English news corpus translated specifically for training NMT systems. The second problem involves the domain-adaptation technique. NMT systems suffer degraded performance when trained with mixed data having different features, such as noisy data and clean data. Though the existing methods try to overcome this problem by using tags for distinguishing the differences between corpora, it is not sufficient. We thus extend a domain-adaptation method using multi-tags to train an NMT model effectively with the clean corpus and existing parallel news corpora with some types of noise. Experimental results show that our corpus increases the translation quality, and that our domain-adaptation method is more effective for learning with the multiple types of corpora than existing domain-adaptation methods are.

In this paper, we focus on developing a Neural Machine Translation (NMT) system on English-to-Traditional-Chinese translation for financial prospectuses of companies which seek listing on the Hong Kong Stock Exchange. To the best of our knowledge, this is the first work on NMT for this specific domain. We propose a domain-specific NMT system by introducing a domain flag to indicate the target-side domain. By training the NMT model on the data from both the IPO corpus and the general domain corpus, we can expand the vocabulary while capturing the common writing styles and sentence structures. Our experimental results show that the proposed NMT system can achieve a significant improvement on translating the IPO documents. More significantly, through a blind assessment by a translator expert, our system outperforms two mainstream commercial tools, the Google translator and SDL Trado for some IPO documents.

State-of-the-art (SOTA) neural machine translation (NMT) systems translate texts at sentence level, ignoring context: intra-textual information, like the previous sentence, and extra-textual information, like the gender of the speaker. As a result, some sentences are translated incorrectly. Personalised NMT (PersNMT) and document-level NMT (DocNMT) incorporate this information into the translation process. Both fields are relatively new and previous work within them is limited. Moreover, there are no readily available robust evaluation metrics for them, which makes it difficult to develop better systems, as well as track global progress and compare different methods. This thesis proposal focuses on PersNMT and DocNMT for the domain of dialogue extracted from TV subtitles in five languages: English, Brazilian Portuguese, German, French and Polish. Three main challenges are addressed: (1) incorporating extra-textual information directly into NMT systems; (2) improving the machine translation of cohesion devices; (3) reliable evaluation for PersNMT and DocNMT.

The task of bilingual dictionary induction (BDI) is commonly used for intrinsic evaluation of cross-lingual word embeddings. The largest dataset for BDI was generated automatically, so its quality is dubious. We study the composition and quality of the test sets for five diverse languages from this dataset, with concerning findings: (1) a quarter of the data consists of proper nouns, which can be hardly indicative of BDI performance, and (2) there are pervasive gaps in the gold-standard targets. These issues appear to affect the ranking between cross-lingual embedding systems on individual languages, and the overall degree to which the systems differ in performance. With proper nouns removed from the data, the margin between the top two systems included in the study grows from 3.4% to 17.2%. Manual verification of the predictions, on the other hand, reveals that gaps in the gold standard targets artificially inflate the margin between the two systems on English to Bulgarian BDI from 0.1% to 6.7%. We thus suggest that future research either avoids drawing conclusions from quantitative results on this BDI dataset, or accompanies such evaluation with rigorous error analysis.

Prior works have demonstrated that a lowresource language pair can benefit from multilingual machine translation (MT) systems, which rely on many language pairs' joint training.This paper proposes two simple strategies to address the rare word issue in multilingual MT systems for two low-resource language pairs: French-Vietnamese and English-Vietnamese.The first strategy is about dynamical learning word similarity of tokens in the shared space among source languages while another one attempts to augment the translation ability of rare words through updating their embeddings during the training.Besides, we leverage monolingual data for multilingual MT systems to increase the amount of synthetic parallel corpora while dealing with the data sparsity problem.We have shown significant improvements of up to +1.62 and +2.54 BLEU points over the bilingual baseline systems for both language pairs and released our datasets for the research community.

International professionals such as Spanish wine and olive oil experts need to write technical texts in English to participate in the international market. To achieve it, many L2 English professionals resort to language technologies to help them obtain domain-specific texts in English. However, available tools are often prone to linguistic and/or domain-specific mistakes. Not only that, but this problem is also aggravated for users of linguistically low-resource language varieties, who have few reliable tools at hand: when the available bilingual data is scarce or defective, it is not easy to develop a language- and domain-compliant writing tool. Such added limitation is also exacerbated by current neural systems’ need of big datasets to achieve SOTA performance. This dissertation proposes a data-centric Corpus Linguistics-informed intervention focused on terminology injection as a domain adaptation strategy for neural bilingual production of low-resource language varieties. On the one hand, a language model based on small comparable domain-specific monolingual corpora is used to select the most similar data that had been automatically downloaded from domain-specific, selected sources. On the other hand, a full-form Spanish-English glossary is employed as a terminological reference to filter or curate big corpora so that language- and domain-adequate equivalences are learned by the system. Additionally, a backtranslation approach is used to augment the datasets used to train the system. To assess the proposed domain adaptation protocol, a set of experimental Neural MT systems were evaluated and compared among them and with commercial system Google Translate from three perspectives: automated metrics, human judgement, and comparison to the gold-standard terminological reference. Results of the automated evaluation of the experimental and commercial systems suggest the domain-adapted systems outperform Google Translate in the translation of wine and olive oil tasting notes. Nevertheless, according to human judgment, the best-scoring experimental system is outperformed by Google Translate in terms of general performance and amount and severity of terminological errors. Most importantly, the findings show an improvement in the terminological performance of the experimental systems after the training using the domain-specific curated data. These results are in line with or surpass previous experimental systems trained on different data-centric domain adaptation strategies based on terminology injection which are usually more computationally demanding and less efficient than the techniques here proposed. However, diversity in evaluation frameworks, language pairs, and lack of detailed results often hinder comparability with previous literature. In sum, and in spite of the many limitations of the study, this dissertation shows the potential of Corpus Linguistics for the development of domain-adapted neural language production tools aimed at aiding Spanish professionals of linguistically low-resource fields successfully engage in international professional communication.