Recently, $k$NN-MT has shown the promising capability of directly incorporating the pre-trained neural machine translation (NMT) model with domain-specific token-level $k$-nearest-neighbor ($k$NN) retrieval to achieve domain adaptation without retraining. Despite being conceptually attractive, it heavily relies on high-quality in-domain parallel corpora, limiting its capability on unsupervised domain adaptation, where in-domain parallel corpora are scarce or nonexistent. In this paper, we propose a novel framework that directly uses in-domain monolingual sentences in the target language to construct an effective datastore for $k$-nearest-neighbor retrieval. To this end, we first introduce an autoencoder task based on the target language, and then insert lightweight adapters into the original NMT model to map the token-level representation of this task to the ideal representation of translation task. Experiments on multi-domain datasets demonstrate that our proposed approach significantly improves the translation accuracy with target-side monolingual data, while achieving comparable performance with back-translation.

Non-autoregressive Transformer is a promising text generation model. However, current non-autoregressive models still fall behind their autoregressive counterparts in translation quality. We attribute this accuracy gap to the lack of dependency modeling among decoder inputs. In this paper, we propose CNAT, which learns implicitly categorical codes as latent variables into the non-autoregressive decoding. The interaction among these categorical codes remedies the missing dependencies and improves the model capacity. Experiment results show that our model achieves comparable or better performance in machine translation tasks, compared with several strong baselines.

Social recommendation is effective in improving the recommendation performance by leveraging social relations from online social networking platforms. Social relations among users provide friends' information for modeling users' interest in candidate items and help items expose to potential consumers (i.e., item attraction). However, there are two issues haven't been well-studied: Firstly, for the user interests, existing methods typically aggregate friends' information contextualized on the candidate item only, and this shallow context-aware aggregation makes them suffer from the limited friends' information. Secondly, for the item attraction, if the item's past consumers are the friends of or have a similar consumption habit to the targeted user, the item may be more attractive to the targeted user, but most existing methods neglect the relation enhanced context-aware item attraction. To address the above issues, we proposed DICER (Dual Side Deep Context-aware Modulation for SocialRecommendation). Specifically, we first proposed a novel graph neural network to model the social relation and collaborative relation, and on top of high-order relations, a dual side deep context-aware modulation is introduced to capture the friends' information and item attraction. Empirical results on two real-world datasets show the effectiveness of the proposed model and further experiments are conducted to help understand how the dual context-aware modulation works.

Predicting clinical outcome is remarkably important but challenging. Research efforts have been paid on seeking significant biomarkers associated with the therapy response or/and patient survival. However, these biomarkers are generally costly and invasive, and possibly dissatifactory for novel therapy. On the other hand, multi-modal, heterogeneous, unaligned temporal data is continuously generated in clinical practice. This paper aims at a unified deep learning approach to predict patient prognosis and therapy response, with easily accessible data, e.g., radiographics, laboratory and clinical information. Prior arts focus on modeling single data modality, or ignore the temporal changes. Importantly, the clinical time series is asynchronous in practice, i.e., recorded with irregular intervals. In this study, we formalize the prognosis modeling as a multi-modal asynchronous time series classification task, and propose a MIA-Prognosis framework with Measurement, Intervention and Assessment (MIA) information to predict therapy response, where a Simple Temporal Attention (SimTA) module is developed to process the asynchronous time series. Experiments on synthetic dataset validate the superiory of SimTA over standard RNN-based approaches. Furthermore, we experiment the proposed method on an in-house, retrospective dataset of real-world non-small cell lung cancer patients under anti-PD-1 immunotherapy. The proposed method achieves promising performance on predicting the immunotherapy response. Notably, our predictive model could further stratify low-risk and high-risk patients in terms of long-term survival.

Cross-prompt automated essay scoring (AES) requires the system to use non target-prompt essays to award scores to a targetprompt essay. Since obtaining a large quantity of pre-graded essays to a particular prompt is often difficult and unrealistic, the task of crossprompt AES is vital for the development of real-world AES systems, yet it remains an under-explored area of research. Models designed for prompt-specific AES rely heavily on prompt-specific knowledge and perform poorly in the cross-prompt setting, whereas current approaches to cross-prompt AES either require a certain quantity of labelled targetprompt essays or require a large quantity of unlabelled target-prompt essays to perform transfer learning in a multi-step manner. To address these issues, we introduce Prompt Agnostic Essay Scorer (PAES) for cross-prompt AES. Our method requires no access to labelled or unlabelled target-prompt data during training and is a single-stage approach. PAES is easy to apply in practice and achieves state-of-the-art performance on the Automated Student Assessment Prize (ASAP) dataset. Keywords: Cross-prompt automated essay scoring · Non prompt-specific features · Neural networks · Domain generalization.

China Abstract Graph neural networks (GNNs) have shown great power in learning on attributed graphs. However, it is still a challenge for GNNs to utilize information faraway from the source node. Moreover, general GNNs require graph attributes as input, so they cannot be appled to plain graphs. In the paper, we propose new models named G-GNNs (Global information for GNNs) to address the above limitations. First, the global structure and attribute features for each node are obtained via unsupervised pre-training, which preserve the global information associated to the node. Then, using the global features and the raw network attributes, we propose a parallel framework of GNNs to learn different aspects from these features. The proposed learning methods can be applied to both plain graphs and attributed graphs. Extensive experiments have shown that G-GNNs can outperform other state-of-the-art models on three standard evaluation graphs. Specially, our methods establish new benchmark records on Cora (84.31%) and Pubmed (80.95%) when learning on attributed graphs. 1 Introduction Semi-supervised learning on graphs aims to recover the labels for all nodes, while only very small proportion of node labels are given. The setting is popular in various of real world applications, since many labels are often expensive and difficult to collect.

Generative adversarial nets (GAN) has been successfully introduced for generating text to alleviate the exposure bias. However, discriminators in these models only evaluate the entire sequence, which causes feedback sparsity and mode collapse. To tackle these problems, we propose a novel mechanism. It first segments the entire sequence into several sub-sequences. Then these sub-sequences, together with the entire sequence, are evaluated individually by the discriminator. At last these feedback signals are all used to guide the learning of GAN. This mechanism learns the generation of both the entire sequence and the sub-sequences simultaneously. Learning to generate sub-sequences is easy and is helpful in generating an entire sequence. It is easy to improve the existing GAN-based models with this mechanism. We rebuild three previous well-designed models with our mechanism, and the experimental results on benchmark data show these models are improved significantly, the best one outperforms the state-of-the-art model.\footnote[1]{All code and data are available at https://github.com/liyzcj/seggan.git

Neural network methods have achieved great success in reviews sentiment classification. Recently, some works achieved improvement by incorporating user and product information to generate a review representation. However, in reviews, we observe that some words or sentences show strong user's preference, and some others tend to indicate product's characteristic. The two kinds of information play different roles in determining the sentiment label of a review. Therefore, it is not reasonable to encode user and product information together into one representation. In this paper, we propose a novel framework to encode user and product information. Firstly, we apply two individual hierarchical neural networks to generate two representations, with user attention or with product attention. Then, we design a combined strategy to make full use of the two representations for training and final prediction. The experimental results show that our model obviously outperforms other state-of-the-art methods on IMDB and Yelp datasets. Through the visualization of attention over words related to user or product, we validate our observation mentioned above.