Language Machine Translation Research Articles

A Novel Natural Language Processing (NLP)–Based Machine Translation Model for English to Pakistan Sign Language Translation

The deaf community in the world uses a gesture-based language, generally known as sign language. Every country has a different sign language; for instance, USA has American Sign Language (ASL) and UK has British Sign Language (BSL). The deaf community in Pakistan uses Pakistan Sign Language (PSL), which like other natural languages, has a vocabulary, sentence structure, and word order. Majority of the hearing community is not aware of PSL due to which there exists a huge communication gap between the two groups. Similarly, deaf persons are unable to read text written in English and Urdu. Hence, the provision of an effective translation model can support the cognitive capability of the deaf community to interpret natural language materials available on the Internet and in other useful resources. This research involves exploiting natural language processing (NLP) techniques to support the deaf community by proposing a novel machine translation model that translates English sentences into equivalent Pakistan Sign Language (PSL). Though a large number of machine translation systems have been successfully implemented for natural to natural language translations, natural to sign language machine translation is a relatively new area of research. State-of-the-art works in natural to sign language translation are mostly domain specific and suffer from low accuracy scores. Major reasons are specialised language structures for sign languages, and lack of annotated corpora to facilitate development of more generalisable machine translation systems. To this end, a grammar-based machine translation model is proposed to translate sentences written in English language into equivalent PSL sentences. To the best of our knowledge, this is a first effort to translate any natural language to PSL using core NLP techniques. The proposed approach involves a structured process to investigate the linguistic structure of PSL and formulate the grammatical structure of PSL sentences. These rules are then formalised into a context-free grammar, which, in turn, can be efficiently implemented as a parsing module for translation and validation of target PSL sentences. The whole concept is implemented as a software system, comprising the NLP pipeline and an external service to render the avatar-based video of translated words, in order to compensate the cognitive hearing deficit of deaf people. The accuracy of the proposed translation model has been evaluated manually and automatically. Quantitative results reveal a very promising Bilingual Evaluation Understudy (BLEU) score of 0.78. Subjective evaluations demonstrate that the system can compensate for the cognitive hearing deficit of end users through the system output expressed as a readily interpretable avatar. Comparative analysis shows that our proposed system works well for simple sentences but struggles to translate compound and compound complex sentences correctly, which warrants future ongoing research.

Predicting Ranking for Scientific Research Papers using Scalable TensorFlow Library and Learning to Rank

Scientific research papers play a vital role for innovation of new technology. It is the future of the development where a novice person can understand the technology and tries to develop a new idea. In this paper, concentrated on relative order for a group of items applied to scientific research paper. In this process we identify how LTR differs from standard supervised learning in the sense that instead of looking at a precise score or class for each sample, it aims to discover the best relative order for a group of items. Firstly we identified the work of ranking of scientific research papers using traditional method know as supervised learning. Secondly we evaluated and made the comparison between the supervised learning and the scalable Tensor flow library for learning to rank. Apart from solving information retrieval problems, Learning to Ranking is mostly used in areas like Natural language processing (NLP), Machine translation, Computational biology or Sentiment analysis.

Improving Self-Attention Networks With Sequential Relations

Recently, self-attention networks show strong advantages of sentence modeling in many NLP tasks. However, self-attention mechanism computes the interactions of every pair of words independently regardless of their positions, which makes it not able to capture the sequential relations between words in different positions in a sentence. In this paper, we improve the self-attention networks by better integrating sequential relations, which is essential for modeling natural languages. Specifically, we 1) propose a position-based attention to model the interaction between two words regarding positions; 2) perform separated attention for the context before and after the current position, respectively; and 3) merge the above two parts with a position-aware gated fusion mechanism. Experiments in natural language inference, machine translation and sentiment analysis tasks show that our sequential relation modeling helps self-attention networks outperform existing approaches. We also provide extensive analyses to shed light on what the models have learned about the sequential relations.

Word Sense Disambiguation in Tamil using Indo Wordnet and Cross-Language Semantic Similarity

Word sense disambiguation is the way to compute the correct sense of a word. It is considered as one of the important subtasks in natural language processing, machine translation and information retrieval. WSD found improving the overall performances of these systems. The job of WSD is to eliminate all senses of a word except the appropriate one as per the given context. The work in Tamil linguistics domain for information retrieval or natural language processing is very less. WSD can be performed in supervised and unsupervised manner. Here, we have proposed an unsupervised approach to disambiguate Tamil words in a given context using the context words and their dictionary gloss definitions. We have proposed two variants of our approach. The first approach uses the number of word overlapping between the glosses of context words whereas the second one uses the similarity between the glosses of context words with that of the ambiguous word. The second one found best among the two. For our approach, we have used Tamil Indo-WordNet, Oxford Tamil Dictionary and English WordNet dictionary glosses. Our method achieves better result in recognising correct senses in Tamil text.

Similarity Based Information Retrieval Using Levenshtein Distance Algorithm

Sentence similarity plays an important role in many text-related research and applications such as information retrieval, information recommendation, natural language processing, machine translation and translation memory, and etc. Calculating similarity between sentences is the basis of measuring the similarity between texts which is the key to document classification and clustering. Sentence similarity partially depends on the word similarity. This system will display a similar text of field, areas and other facts in document retrieval. This paper uses a sentence matching method of the Levenshtein Distance algorithm. Thesimilarity between words can be calculated from the spelling of words or the meaning of words. Sentence similarity: The similarities between words in different sentences have a great influence on the similarity between two sentences. This system is retrieved from a similar sentence that included Theories, Methods and other facts in the document database.

Support of Arabic Sign Language Machine Translation based on Morphological processing

Support of Arabic Sign Language Machine Translation based on Morphological processing

An Automatic and a Machine-assisted Method to Clean Bilingual Corpus

Two different methods of corpus cleaning are presented in this article. One is a machine-assisted technique, which is good to clean small-sized parallel corpus, and the other is an automatic method, which is suitable for cleaning large-sized parallel corpus. A baseline SMT (MOSES) system is used to evaluate these methods. The machine-assisted technique used two features: word alignment and length of the source and target language sentence. These features are used to detect mistranslations in the corpus, which are then handled by a human translator. Experiments of this method are conducted on the English-to-Indian Language Machine Translation (EILMT) corpus (English-Hindi). The Bilingual Evaluation Understudy (BLEU) score is improved by 0.47% for the clean corpus. Automatic method of corpus cleaning uses a combination of two features. One feature is length of source and target language sentence and the second feature is Viterbi alignment score generated by Hidden Markov Model for each sentence pair. Two different threshold values are used for these two features. These values are decided by using a small-sized manually annotated parallel corpus of 206 sentence pairs. Experiments of this method are conducted on the HindEnCorp corpus, released in the workshop of the Association of Computational Linguistics (ACL 2014). The BLEU score is improved by 0.6% on clean corpus. A comparison of the two methods is also presented on EILMT corpus.

Study on text representation method based on deep learning and topic information

Deep learning provides a new modeling method for natural language processing. In recent years, it has been applied in language model, text classification, machine translation, sentiment analysis, question and answer system, word distributed representation, etc., and a series of theoretical research results have been obtained. For the text representation task, this paper studies the strategy of fusing global and local context information, and proposes a word representation model called Topic-based CBOW that integrates deep neural network, topic information and word order information. Then, based on the word distributed representation obtained by Topic-based CBOW, a short text representation method with TF–IWF-weighted pooling is proposed. Finally, the performance of the Topic-based CBOW model and the short text representation are compared with the baseline models, and it is found that the proposed method improves the quality of the word distributed representation to some extent by introducing the topic vector and retaining word order information, and text representation also performs well in text classification tasks.

Indowordnet’s help in Indian language machine translation

Languages with insufficient digitally available resources, such as, Indian–Indian and English–Indian language Machine Translation (MT) system developments, faces the difficulty to translate various lexical phenomena. In this paper, we present our work on a comparative study of 440 phrase-based statistical trained models for 110 language pairs across 11 Indian languages. We have developed 110 baseline statistical machine translation systems. Then, we have augmented the training corpus with Indowordnet synset word entries of lexical database and further trained 110 models on top of the baseline system. We have done a detailed performance comparison using various evaluation metrics such as BLEU score, METEOR, and TER. We observed significant improvement in evaluations of translation quality across all the 440 models after using the Indowordnet. These experiments give a detailed insight in two ways: (1) usage of lexical database with synset mapping for resource poor languages and (2) efficient usage of Indowordnet synset mapping. Moreover, synset mapped lexical entries helped the SMT system to handle the ambiguity to a great extent during the translation.

English text to Indian Sign Language Machine Translation: A Rule Based Method

The hearing challenged community all over world face difficulties to communicate with others. Machine translation has been one of the prominent technologies to facilitate a two way communication to the deaf and hard of hearing community all over the world. We have explored and formulated the fundamental rules of Indian Sign Language and implemented as a translation mechanism of English Text to Indian sign Language glosses. The structure of the source text is identified and transferred to the target language according to the formulated rules and sub rules. The intermediate phases of the transfer process is also mentioned in this research work.

Towards an open platform for machine translation of spoken languages into sign languages

The purpose of this paper is to investigate the feasibility of offering a multilingual platform for text-to-sign translation, i.e., a solution where a machine translates digital contents in several spoken languages to several sign languages in scenarios such as Digital TV, Web and Cinema. This solution—called OpenSigns—is an open platform that has several common components for generic functionalities originating from the Suite VLibras, including the creation and manipulation of 3D animation models, and interchangeable mechanisms specific for each sign language, such as a text-to-gloss machine translation engine and a signs dictionary for each sign language. Our motivation is that the concentration of efforts and resources around a single solution could provide some state-of-the-art improvement, such as a standard solution for the industry and a greater functional flexibility for common components. In addition, we could share techniques and heuristics between the translation mechanisms, reducing the effort to make a new sign language available on the platform, which may further enhance digital inclusion and accessibility, especially for the poorest countries.

RNN-LSTM-GRU based language transformation

In past, rule-based and statistical machine translation techniques were employed to solve Urdu transliteration techniques. As mentioned in the literature, Urdu is considered as low-resource language. An impressive effort has been made for Arabic, French and Chinese language transliteration as compared to the Urdu language. Machine translation of Urdu language is a challenging problem. A very minute research work has been conducted toward Urdu transliteration. Factors behind the ignorance of Urdu language in research may be for its morphological complexity, diversity and most importantly due to the lack of reasonable bilingual parallel dataset. Getting a corpus for a language transliteration is the main resource to work on. This paper demonstrates the application of neural machine translation (NMT) for Urdu language transliteration, with the emphasis on contextual coverage of a language, which helps to improve transliteration accuracy. Build a robust NMT model which delivers efficient performance when trained over bilingual parallel corpora. Neural machine translation is an emerging technique depicting impressive performance, better than traditional MT methods in multiple aspects. In this research, we build the NMT model for the Urdu language to improve transliteration quality. An attention-based encoder–decoder system is proposed, and our experiment proves the efficiency of the proposed approach. To the best of our knowledge, this is the first effort for Urdu language bidirectional transliteration toward neural machine translation.

Machine Translation in Indian Languages: Challenges and Resolution

Abstract English to Indian language machine translation poses the challenge of structural and morphological divergence. This paper describes English to Indian language statistical machine translation using preordering and suffix separation. The preordering uses rules to transfer the structure of the source sentences prior to training and translation. This syntactic restructuring helps statistical machine translation to tackle the structural divergence and hence provides better translation quality. The suffix separation is used to tackle the morphological divergence between English and highly agglutinative Indian languages. We demonstrate that the use of preordering and suffix separation helps in improving the quality of English to Indian language machine translation.

Special issue on “Multi‐modal information learning and analytics of cross‐media big data”

Special issue on “Multi‐modal information learning and analytics of cross‐media big data”

Research on Neural Machine Translation Model

In neural machine translation (NMT), cyclic neural networks, especially long-term and short-term memory networks and gated recurrent neural networks, have been regarded as the latest methods for sequence modeling and transduction problems for a long time, such as language modeling and machine translation. When the cyclic neural network is running, the sequence information is processed one by one, strictly following the order from left to right or from right to left, processing one word at a time, and parallel operation cannot be realized, resulting in slow running speed. With the rapid development of neural machine translation (NMT) network architecture, cyclic neural network has been effectively replaced by convolution network and self - attention. Convolution neural network has replaced the divine circulation neural network due to its parallel computation of convolution. The Transformer model replaces the long-term and short-term memory network with a complete self-attention structure, and abandons the traditional encoder and decoder model which must combine the inherent mode of convolutional neural network or circular neural network and only uses the self-attention mechanism. Although the biggest innovation of Transformer architecture is to use full self - attention, there are several other factors, such as multi-head attention and residual connection. The model flexibly combine several common building blocks in the Transformer architecture with the cyclic neural network. By borrowing the framework of the Transformer architecture without using full self - attention, experiments show that the cyclic model can be very close to the performance of the Transformer Our model achieved 26.7 BLEU in the WMT 2014 English to German translation task and 37.8 BLEU in the WMT 2014 English to French translation task. Using these two scores alone is very close to the score of the Transformer architecture using full attention, so even if the cyclic neural network is used instead of full self - attention, it can perform well on the data set.

Apps-based Machine Translation on Smart Media Devices - A Review

Machine Translation Systems are part of Natural Language Processing (NLP) that makes communication possible among people using their own native language through computer and smart media devices. This paper describes recent progress in language dictionaries and machine translation commonly used for communications and social interaction among people or Internet users worldwide who speak different languages. Problems of accuracy and quality related to computer translation systems encountered in web & Apps-based translation are described and discussed. Possible programming solutions to the problems are also put forward to create software tools that are able to analyze and synthesize language intelligently based on semantic representation of sentences and phrases. Challenges and problems on Apps-based machine translation on smart devices towards AI, NLP, smart learning and understanding still remain until now, and need to be addressed and solved through collaboration between computational linguists and computer scientists.

Neural machine translation system for the Kazakh language based on synthetic corpora

The lack of big parallel data is present for the Kazakh language. This problem seriously impairs the quality of machine translation from and into Kazakh. This article considers the neural machine translation of the Kazakh language on the basis of synthetic corpora. The Kazakh language belongs to the Turkic languages, which are characterised by rich morphology. Neural machine translation of natural languages requires large training data. The article will show the model for the creation of synthetic corpora, namely the generation of sentences based on complete suffixes for the Kazakh language. The novelty of this approach of the synthetic corpora generation for the Kazakh language is the generation of sentences on the basis of the complete system of suffixes of the Kazakh language. By using generated synthetic corpora we are improving the translation quality in neural machine translation of Kazakh-English and Kazakh-Russian pairs.

A Cloud Server Oriented FPGA Accelerator for LSTM Recurrent Neural Network

Long Short-Term Memory network (LSTM) is the most widely used recurrent neural network architecture. It plays an important role in a number of research areas, such as language modeling, machine translation and image captioning. However, owing to its recurrent nature, general-purpose processors like CPUs and GPGPUs can achieve limited parallelism while consuming high power energy. FPGA accelerators can outperform general-purpose processors with flexibility, energy-efficiency and more delicate optimization capabilities for the recurrent based algorithms. In this paper, we present the design and implementation of a cloud-oriented FPGA accelerator for LSTM. Different from most of previous works designed for embedded systems, our FPGA accelerator transfers data sequences from and to the host server through PCIe and performs multiple time series predictions in parallel. We optimize both the on-chip computation and the communication between the host server and the FPGA board. We perform experiments to evaluate the overall performance as well as the computation and the PCIe communication efforts. The results show that the performance of our implementation is better than the CPU-based and other hardware-based implementations.

The parallel corpus for information extraction based on natural language processing and machine translation

AbstractIn conventional algorithms, the lack of entity information, reference, and semantic relations in the current corpus leads to a low rate of precision and efficiency in constructing cross‐language bilingual mapping. According to natural language processing and machine translation technology, to solve the problem, this paper aims to establish a parallel corpus for information extraction based on the OntoNotes corpus by combining automatic extraction and manual adjustment. To verify the validity of the parallel corpus constructed in this paper, a comparative experiment was carried out on the corpus. The corpus entity alignment rate, anaphora absence, and syntactic structure were analysed in detail based on statistics. The data set is well performed in language processing and machine translation. The parallel corpus for information extraction constructed in this paper can produce highly precise, stable, and efficient information in the process of bilingual mapping, which provides an effective parallel corpus for the study in machine translation of bilingual mapping.

Ethical issues regarding machine(-assisted) translation of literary texts

ABSTRACTThis article investigates essential ethical issues that should be taken into consideration when adopting or tailoring technological tools for literary translation. The discussion on ethical issues draws on recent studies on translation technology and on the usage of machine(-assisted) translation for literary language. An overview of the consequences of the recent increase in technologization for both non-literary and literary translation is provided and an argument for sustainable development in literary translation is made, based on a holistic understanding of translation quality. The notion of voice is taken as an example of the special challenges related to the translation of literary language, which research on machine(-assisted) translation of literary language has not yet sufficiently taken into account. Lastly, avant-garde aesthetic views and usages of machine translation are presented through the notion of noise.