PyTorchKR Papers

최신 논문

총 5,206편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1412.6650

Incremental Adaptation Strategies for Neural Network Language Models

It is today acknowledged that neural network language models outperform backoff language models in applications like speech recognition or statistical machine translation. However, training these models on large amounts of data can take several days. We present efficient techniques to adapt a neural network language model to new data. Instead of training a completely new model or relying on mixture approaches, we propose two new methods: continued training on resampled data or insertion of adaptation layers. We present experimental results in an CAT environment where the post-edits of professional translators are used to improve an SMT system. Both methods are very fast and achieve significant improvements without overfitting the small adaptation data.

arXiv:1507.03223

Classifier-Based Text Simplification for Improved Machine Translation

Machine Translation is one of the research fields of Computational Linguistics. The objective of many MT Researchers is to develop an MT System that produce good quality and high accuracy output translations and which also covers maximum language pairs. As internet and Globalization is increasing day by day, we need a way that improves the quality of translation. For this reason, we have developed a Classifier based Text Simplification Model for English-Hindi Machine Translation Systems. We have used support vector machines and Na\"ive Bayes Classifier to develop this model. We have also evaluated the performance of these classifiers.

arXiv:1507.07826

Classifying informative and imaginative prose using complex networks

Statistical methods have been widely employed in recent years to grasp many language properties. The application of such techniques have allowed an improvement of several linguistic applications, which encompasses machine translation, automatic summarization and document classification. In the latter, many approaches have emphasized the semantical content of texts, as it is the case of bag-of-word language models. This approach has certainly yielded reasonable performance. However, some potential features such as the structural organization of texts have been used only on a few studies. In this context, we probe how features derived from textual structure analysis can be effectively employed in a classification task. More specifically, we performed a supervised classification aiming at discriminating informative from imaginative documents. Using a networked model that describes the local topological/dynamical properties of function words, we achieved an accuracy rate of up to 95%, which is much higher than similar networked approaches. A systematic analysis of feature relevance revealed that symmetry and accessibility measurements are among the most prominent network measurements. Our results suggest that these measurements could be used in related language applications, as they play a complementary role in characterizing texts.

arXiv:1508.02445

Removing Biases from Trainable MT Metrics by Using Self-Training

Most trainable machine translation (MT) metrics train their weights on human judgments of state-of-the-art MT systems outputs. This makes trainable metrics biases in many ways. One of them is preferring longer translations. These biased metrics when used for tuning are evaluating different types of translations -- n-best lists of translations with very diverse quality. Systems tuned with these metrics tend to produce overly long translations that are preferred by the metric but not by humans. This is usually solved by manually tweaking metric's weights to equally value recall and precision. Our solution is more general: (1) it does not address only the recall bias but also all other biases that might be present in the data and (2) it does not require any knowledge of the types of features used which is useful in cases when manual tuning of metric's weights is not possible. This is accomplished by self-training on unlabeled n-best lists by using metric that was initially trained on standard human judgments. One way of looking at this is as domain adaptation from the domain of state-of-the-art MT translations to diverse n-best list translations.

arXiv:1508.02285

Adapting Phrase-based Machine Translation to Normalise Medical Terms in Social Media Messages

Previous studies have shown that health reports in social media, such as DailyStrength and Twitter, have potential for monitoring health conditions (e.g. adverse drug reactions, infectious diseases) in particular communities. However, in order for a machine to understand and make inferences on these health conditions, the ability to recognise when laymen's terms refer to a particular medical concept (i.e.\ text normalisation) is required. To achieve this, we propose to adapt an existing phrase-based machine translation (MT) technique and a vector representation of words to map between a social media phrase and a medical concept. We evaluate our proposed approach using a collection of phrases from tweets related to adverse drug reactions. Our experimental results show that the combination of a phrase-based MT technique and the similarity between word vector representations outperforms the baselines that apply only either of them by up to 55%.

arXiv:1508.02142

Feature-based Decipherment for Large Vocabulary Machine Translation

Orthographic similarities across languages provide a strong signal for probabilistic decipherment, especially for closely related language pairs. The existing decipherment models, however, are not well-suited for exploiting these orthographic similarities. We propose a log-linear model with latent variables that incorporates orthographic similarity features. Maximum likelihood training is computationally expensive for the proposed log-linear model. To address this challenge, we perform approximate inference via MCMC sampling and contrastive divergence. Our results show that the proposed log-linear model with contrastive divergence scales to large vocabularies and outperforms the existing generative decipherment models by exploiting the orthographic features.

arXiv:1508.05051

Auto-Sizing Neural Networks: With Applications to n-gram Language Models

Neural networks have been shown to improve performance across a range of natural-language tasks. However, designing and training them can be complicated. Frequently, researchers resort to repeated experimentation to pick optimal settings. In this paper, we address the issue of choosing the correct number of units in hidden layers. We introduce a method for automatically adjusting network size by pruning out hidden units through $\ell_{\infty,1}$ and $\ell_{2,1}$ regularization. We apply this method to language modeling and demonstrate its ability to correctly choose the number of hidden units while maintaining perplexity. We also include these models in a machine translation decoder and show that these smaller neural models maintain the significant improvements of their unpruned versions.

arXiv:1508.04271

Probabilistic Modelling of Morphologically Rich Languages

This thesis investigates how the sub-structure of words can be accounted for in probabilistic models of language. Such models play an important role in natural language processing tasks such as translation or speech recognition, but often rely on the simplistic assumption that words are opaque symbols. This assumption does not fit morphologically complex language well, where words can have rich internal structure and sub-word elements are shared across distinct word forms. Our approach is to encode basic notions of morphology into the assumptions of three different types of language models, with the intention that leveraging shared sub-word structure can improve model performance and help overcome data sparsity that arises from morphological processes. In the context of n-gram language modelling, we formulate a new Bayesian model that relies on the decomposition of compound words to attain better smoothing, and we develop a new distributed language model that learns vector representations of morphemes and leverages them to link together morphologically related words. In both cases, we show that accounting for word sub-structure improves the models' intrinsic performance and provides benefits when applied to other tasks, including machine translation. We then shift the focus beyond the modelling of word sequences and consider models that automatically learn what the sub-word elements of a given language are, given an unannotated list of words. We formulate a novel model that can learn discontiguous morphemes in addition to the more conventional contiguous morphemes that most previous models are limited to. This approach is demonstrated on Semitic languages, and we find that modelling discontiguous sub-word structures leads to improvements in the task of segmenting words into their contiguous morphemes.

arXiv:1506.00196

Sequence-to-Sequence Neural Net Models for Grapheme-to-Phoneme Conversion

Sequence-to-sequence translation methods based on generation with a side-conditioned language model have recently shown promising results in several tasks. In machine translation, models conditioned on source side words have been used to produce target-language text, and in image captioning, models conditioned images have been used to generate caption text. Past work with this approach has focused on large vocabulary tasks, and measured quality in terms of BLEU. In this paper, we explore the applicability of such models to the qualitatively different grapheme-to-phoneme task. Here, the input and output side vocabularies are small, plain n-gram models do well, and credit is only given when the output is exactly correct. We find that the simple side-conditioned generation approach is able to rival the state-of-the-art, and we are able to significantly advance the stat-of-the-art with bi-directional long short-term memory (LSTM) neural networks that use the same alignment information that is used in conventional approaches.

arXiv:1508.03790

Depth-Gated LSTM

In this short note, we present an extension of long short-term memory (LSTM) neural networks to using a depth gate to connect memory cells of adjacent layers. Doing so introduces a linear dependence between lower and upper layer recurrent units. Importantly, the linear dependence is gated through a gating function, which we call depth gate. This gate is a function of the lower layer memory cell, the input to and the past memory cell of this layer. We conducted experiments and verified that this new architecture of LSTMs was able to improve machine translation and language modeling performances.

arXiv:1509.01938

Exploiting Out-of-Domain Data Sources for Dialectal Arabic Statistical Machine Translation

Statistical machine translation for dialectal Arabic is characterized by a lack of data since data acquisition involves the transcription and translation of spoken language. In this study we develop techniques for extracting parallel data for one particular dialect of Arabic (Iraqi Arabic) from out-of-domain corpora in different dialects of Arabic or in Modern Standard Arabic. We compare two different data selection strategies (cross-entropy based and submodular selection) and demonstrate that a very small but highly targeted amount of found data can improve the performance of a baseline machine translation system. We furthermore report on preliminary experiments on using automatically translated speech data as additional training data.

arXiv:1509.04473

Splitting Compounds by Semantic Analogy

Compounding is a highly productive word-formation process in some languages that is often problematic for natural language processing applications. In this paper, we investigate whether distributional semantics in the form of word embeddings can enable a deeper, i.e., more knowledge-rich, processing of compounds than the standard string-based methods. We present an unsupervised approach that exploits regularities in the semantic vector space (based on analogies such as "bookshop is to shop as bookshelf is to shelf") to produce compound analyses of high quality. A subsequent compound splitting algorithm based on these analyses is highly effective, particularly for ambiguous compounds. German to English machine translation experiments show that this semantic analogy-based compound splitter leads to better translations than a commonly used frequency-based method.

arXiv:1509.03870

The USFD Spoken Language Translation System for IWSLT 2014

The University of Sheffield (USFD) participated in the International Workshop for Spoken Language Translation (IWSLT) in 2014. In this paper, we will introduce the USFD SLT system for IWSLT. Automatic speech recognition (ASR) is achieved by two multi-pass deep neural network systems with adaptation and rescoring techniques. Machine translation (MT) is achieved by a phrase-based system. The USFD primary system incorporates state-of-the-art ASR and MT techniques and gives a BLEU score of 23.45 and 14.75 on the English-to-French and English-to-German speech-to-text translation task with the IWSLT 2014 data. The USFD contrastive systems explore the integration of ASR and MT by using a quality estimation system to rescore the ASR outputs, optimising towards better translation. This gives a further 0.54 and 0.26 BLEU improvement respectively on the IWSLT 2012 and 2014 evaluation data.

arXiv:1509.05517

A Light Sliding-Window Part-of-Speech Tagger for the Apertium Free/Open-Source Machine Translation Platform

This paper describes a free/open-source implementation of the light sliding-window (LSW) part-of-speech tagger for the Apertium free/open-source machine translation platform. Firstly, the mechanism and training process of the tagger are reviewed, and a new method for incorporating linguistic rules is proposed. Secondly, experiments are conducted to compare the performances of the tagger under different window settings, with or without Apertium-style "forbid" rules, with or without Constraint Grammar, and also with respect to the traditional HMM tagger in Apertium.

arXiv:1508.04025

Effective Approaches to Attention-based Neural Machine Translation

An attentional mechanism has lately been used to improve neural machine translation (NMT) by selectively focusing on parts of the source sentence during translation. However, there has been little work exploring useful architectures for attention-based NMT. This paper examines two simple and effective classes of attentional mechanism: a global approach which always attends to all source words and a local one that only looks at a subset of source words at a time. We demonstrate the effectiveness of both approaches over the WMT translation tasks between English and German in both directions. With local attention, we achieve a significant gain of 5.0 BLEU points over non-attentional systems which already incorporate known techniques such as dropout. Our ensemble model using different attention architectures has established a new state-of-the-art result in the WMT'15 English to German translation task with 25.9 BLEU points, an improvement of 1.0 BLEU points over the existing best system backed by NMT and an n-gram reranker.

arXiv:1510.00001

Polish to English Statistical Machine Translation

This research explores the effects of various training settings on a Polish to English Statistical Machine Translation system for spoken language. Various elements of the TED, Europarl, and OPUS parallel text corpora were used as the basis for training of language models, for development, tuning and testing of the translation system. The BLEU, NIST, METEOR and TER metrics were used to evaluate the effects of the data preparations on the translation results.

arXiv:1509.09097

Polish - English Speech Statistical Machine Translation Systems for the IWSLT 2013

This research explores the effects of various training settings from Polish to English Statistical Machine Translation system for spoken language. Various elements of the TED parallel text corpora for the IWSLT 2013 evaluation campaign were used as the basis for training of language models, and for development, tuning and testing of the translation system. The BLEU, NIST, METEOR and TER metrics were used to evaluate the effects of data preparations on translation results. Our experiments included systems, which use stems and morphological information on Polish words. We also conducted a deep analysis of provided Polish data as preparatory work for the automatic data correction and cleaning phase.

arXiv:1509.09093

A Sentence Meaning Based Alignment Method for Parallel Text Corpora Preparation

Text alignment is crucial to the accuracy of Machine Translation (MT) systems, some NLP tools or any other text processing tasks requiring bilingual data. This research proposes a language independent sentence alignment approach based on Polish (not position-sensitive language) to English experiments. This alignment approach was developed on the TED Talks corpus, but can be used for any text domain or language pair. The proposed approach implements various heuristics for sentence recognition. Some of them value synonyms and semantic text structure analysis as a part of additional information. Minimization of data loss was ensured. The solution is compared to other sentence alignment implementations. Also an improvement in MT system score with text processed with described tool is shown.

arXiv:1509.09090

Real-Time Statistical Speech Translation

This research investigates the Statistical Machine Translation approaches to translate speech in real time automatically. Such systems can be used in a pipeline with speech recognition and synthesis software in order to produce a real-time voice communication system between foreigners. We obtained three main data sets from spoken proceedings that represent three different types of human speech. TED, Europarl, and OPUS parallel text corpora were used as the basis for training of language models, for developmental tuning and testing of the translation system. We also conducted experiments involving part of speech tagging, compound splitting, linear language model interpolation, TrueCasing and morphosyntactic analysis. We evaluated the effects of variety of data preparations on the translation results using the BLEU, NIST, METEOR and TER metrics and tried to give answer which metric is most suitable for PL-EN language pair.

arXiv:1509.09088

Enhanced Bilingual Evaluation Understudy

Our research extends the Bilingual Evaluation Understudy (BLEU) evaluation technique for statistical machine translation to make it more adjustable and robust. We intend to adapt it to resemble human evaluation more. We perform experiments to evaluate the performance of our technique against the primary existing evaluation methods. We describe and show the improvements it makes over existing methods as well as correlation to them. When human translators translate a text, they often use synonyms, different word orders or style, and other similar variations. We propose an SMT evaluation technique that enhances the BLEU metric to consider variations such as those.