This paper proposes an approach for applying GANs to NMT. We build a conditional sequence generative adversarial net which comprises of two adversarial sub models, a generator and a discriminator. The generator aims to generate sentences which are hard to be discriminated from human-translated sentences ( i.e., the golden target sentences); And the discriminator makes efforts to discriminate the machine-generated sentences from humantranslated ones. The two sub models play a mini-max game and achieve the win-win situation when they reach a Nash Equilibrium. Additionally, the static sentence-level BLEU is utilized as the reinforced objective for the generator, which biases the generation towards high BLEU points. During training, both the dynamic discriminator and the static BLEU objective are employed to evaluate the generated sentences and feedback the evaluations to guide the learning of the generator. Experimental results show that the proposed model consistently outperforms the traditional RNNSearch and the newly emerged state-ofthe-art Transformer on English-German and Chinese-English translation tasks.

Unsupervised neural machine translation (NMT) is a recently proposed approach for machine translation which aims to train the model without using any labeled data. The models proposed for unsupervised NMT often use only one shared encoder to map the pairs of sentences from different languages to a shared-latent space, which is weak in keeping the unique and internal characteristics of each language, such as the style, terminology, and sentence structure. To address this issue, we introduce an extension by utilizing two independent encoders but sharing some partial weights which are responsible for extracting high-level representations of the input sentences. Besides, two different generative adversarial networks (GANs), namely the local GAN and global GAN, are proposed to enhance the cross-language translation. With this new approach, we achieve significant improvements on English-German, English-French and Chinese-to-English translation tasks.

Aim: Physcion is a major bioactive ingredient in the traditional Chinese medicine Radix et Rhizoma Rhei, which has an anthraquinone chemical structure and exhibits a variety of pharmacological activities including laxative, hepatoprotective, anti-inflammatory, antimicrobial and anti-proliferative effects. In this study we investigated the effect of physcion on human nasopharyngeal carcinoma in vitro and in vivo, as well as the mechanisms underlying the anti-tumor action. Methods: The nasopharyngeal carcinoma cell line CNE2 was treated with physcion, and cell viability was detected using MTT and colony formation assays. Flow cytometry was used to assess the cell cycle arrest, mitochondrial membrane potential loss, apoptosis, autophagy and intracellular ROS generation. Apoptotic cell death was also confirmed by a TUNEL assay. The expression of target or marker molecules was determined using Western blotting. The activity of caspase-3, 8, and 9 was detected with an ELISA kit. A xenograft murine model was used to evaluate the in vivo anti-tumor action of physcion, the mice were administered physcion (10, 20 mg·kg -1 ·d -1 , ip) for 30 d.Results: Treatment with physcion (5, 10, and 20 μmol/L) dose-dependently suppressed the cell viability and colony formation in CNE2 cells. Physcion (10 and 20 μmol/L) dose-dependently blocked cell cycle progression at G 1 phase and induced both caspase-dependent apoptosis and autophagy in CNE2 cells. Furthermore, physcion treatment induced excessive ROS generation in CNE2 cells, and subsequently disrupted the miR-27a/ZBTB10 axis, resulting in repression of the transcription factor Sp1 that was involved in physcioninduced apoptosis and autophagy. Moreover, physcion-induced autophagy acted as a pro-apoptotic factor, and possibly contributed to physcion-induced apoptosis. In the xenograft murine model, administration of physcion dose-dependently suppressed the tumor growth without affecting the body weight. Furthermore, the anti-tumor effects of physcion were correlated with downregulation of Sp1 and suppression of miR-27a in the tumor tissues. Conclusion: Physcion induces apoptosis and autophagy in human nasopharyngeal carcinoma by targeting Sp1, which was mediated by ROS/miR-27a/ZBTB10 signaling. The results suggest that physcion is a promising candidate for the treatment of human nasopharyngeal carcinoma.

No abstract

This paper proposes a new pre-training method, called Code-Switching Pre-training (CSP for short) for Neural Machine Translation (NMT). Unlike traditional pre-training method which randomly masks some fragments of the input sentence, the proposed CSP randomly replaces some words in the source sentence with their translation words in the target language. Specifically, we firstly perform lexicon induction with unsupervised word embedding mapping between the source and target languages, and then randomly replace some words in the input sentence with their translation words according to the extracted translation lexicons. CSP adopts the encoderdecoder framework: its encoder takes the codemixed sentence as input, and its decoder predicts the replaced fragment of the input sentence. In this way, CSP is able to pre-train the NMT model by explicitly making the most of the cross-lingual alignment information extracted from the source and target monolingual corpus. Additionally, we relieve the pretrainfinetune discrepancy caused by the artificial symbols like [mask]. To verify the effectiveness of the proposed method, we conduct extensive experiments on unsupervised and supervised NMT. Experimental results show that CSP achieves significant improvements over baselines without pre-training or with other pre-training methods.