Generative adversarial networks (GAN) have great successes on natural language processing (NLP) and neural machine translation (NMT). However, the existing discriminator in GAN for NMT only combines two words as one query to train the translation models, which restrict the discriminator to be more meaningful and fail to apply rich monolingual information. Recent studies only consider one single reference translation during model training, this limit the GAN model to learn sufficient information about the representation of source sentence. These situations are even worse when languages are morphologically rich. In this article, an extended version of GAN model for neural machine translation is proposed to optimize the performance of morphologically rich language translation. In particular, we use the morphological word embedding instead of word embedding as input in GAN model to enrich the representation of words and overcome the data sparsity problem during model training. Moreover, multiple references are integrated into discriminator to make the model consider more context information and adapt to the diversity of different languages. Experimental results on German $\leftrightarrow$ English, French $\leftrightarrow$ English, Czech $\leftrightarrow$ English, Finnish $\leftrightarrow$ English, Turkish $\leftrightarrow$ English, Chinese $\leftrightarrow$ English, Finnish $\leftrightarrow$ Turkish and Turkish $\leftrightarrow$ Czech translation tasks demonstrate that our method achieves significant improvements over baseline systems.

Bilingual resources play a very important role in many natural language processing tasks, especially the tasks in cross-lingual scenarios. However, it is expensive and time consuming to build such resources. Lexical borrowing happens in almost every language. This inspires us to detect these loanwords effectively, and to use the “loanword (in receipt language)”-“donor word (in donor language)” to extend the bilingual resource for NLP tasks in low-resource languages. In this article, we propose a novel method to identify loanwords in Uyghur. The most important advantage of this method is that the model only relies on large amount of monolingual corpora and only a small scale of annotated data. Our loanword identification model includes two parts: loanword candidate generation and loanword prediction. In the first part, we use two large-scale monolingual corpora and a small bilingual dictionary to train a cross-lingual embedding model. Since semantic unrelated words often cannot be treated as loanword pairs, a loanword candidate list will be generated according to this model and a word list in Uyghur. In the second part, we predict from the preceding candidates based on a log-linear model that integrates several features such as pronunciation similarity, part-of-speech tags, and hybrid language modeling. To evaluate the effectiveness of our proposed method, we conduct two types of experiments: loanword identification and OOV translation. Experimental results showed that (1) our proposed method achieved significant F1 improvements compared to other models in all four loanword identification tasks in Uyghur, and (2) after extending the existing translation models with loanword identification results, OOV rates in several language pairs reduced significantly and the translation performance improved.

As the rapid development of deep learning methods, neural machine translation (NMT) has attracted more and more attention in recent years. However, lack of bilingual resources decreases the performance of the low-resource NMT model seriously. To overcome this problem, several studies put their efforts on knowledge transfer from high-resource language pairs to low-resource language pairs. However, these methods usually focus on one single granularity of language and the parameter sharing among different granularities in NMT is not well studied. In this article, we propose to improve the parameter sharing in low-resource NMT by introducing multi-granularity knowledge such as word, phrase and sentence. This knowledge can be monolingual and bilingual. We build the knowledge sharing model for low-resource NMT based on a multi-task learning framework, three auxiliary tasks such as syntax parsing, cross-lingual named entity recognition, and natural language generation are selected for the low-resource NMT. Experimental results show that the proposed method consistently outperforms six strong baseline systems on several low-resource language pairs.