We present a direct speech-to-speech translation (S2ST) model that translates speech from one language to speech in another language without relying on intermediate text generation. We tackle the problem by first applying a self-supervised discrete speech encoder on the target speech and then training a sequence-to-sequence speech-to-unit translation (S2UT) model to predict the discrete representations of the target speech. When target text transcripts are available, we design a joint speech and text training framework that enables the model to generate dual modality output (speech and text) simultaneously in the same inference pass. Experiments on the Fisher Spanish-English dataset show that the proposed framework yields improvement of 6.7 BLEU compared with a baseline direct S2ST model that predicts spectrogram features. When trained without any text transcripts, our model performance is comparable to models that predict spectrograms and are trained with text supervision, showing the potential of our system for translation between unwritten languages.

The pivot for the unified Aspect-based Sentiment Analysis (ABSA) is to couple aspect terms with their corresponding opinion terms, which might further derive easier sentiment predictions. In this paper, we investigate the unified ABSA task from the perspective of Machine Reading Comprehension (MRC) by observing that the aspect and the opinion terms can serve as the query and answer in MRC interchangeably. We propose a new paradigm named Role Flipped Machine Reading Comprehension (RF-MRC) to resolve. At its heart, the predicted results of either the Aspect Term Extraction (ATE) or the Opinion Terms Extraction (OTE) are regarded as the queries, respectively, and the matched opinion or aspect terms are considered as answers. The queries and answers can be flipped for multi-hop detection. Finally, every matched aspect-opinion pair is predicted by the sentiment classifier. RF-MRC can solve the ABSA task without any additional data annotation or transformation. Experiments on three widely used benchmarks and a challenging dataset demonstrate the superiority of the proposed framework.

A long-standing issue with paraphrase generation is the lack of reliable supervision signals. In this paper, we propose a new unsupervised paradigm for paraphrase generation based on the assumption that the probabilities of generating two sentences with the same meaning given the same context should be the same. Inspired by this fundamental idea, we propose a pipelined system which consists of paraphrase candidate generation based on contextual language models, candidate filtering using scoring functions, and paraphrase model training based on the selected candidates. The proposed paradigm offers merits over existing paraphrase generation methods: (1) using the context regularizer on meanings, the model is able to generate massive amounts of high-quality paraphrase pairs; (2) the combination of the huge amount of paraphrase candidates and further diversity-promoting filtering yields paraphrases with more lexical and syntactic diversity; and (3) using human-interpretable scoring functions to select paraphrase pairs from candidates, the proposed framework provides a channel for developers to intervene with the data generation process, leading to a more controllable model. Experimental results across different tasks and datasets demonstrate that the proposed paradigm significantly outperforms existing paraphrase approaches in both supervised and unsupervised setups.

In this paper, we formulate the personalized news headline generation problem whose goal is to output a user-specific title based on both a user’s reading interests and a candidate news body to be exposed to her. To build up a benchmark for this problem, we publicize a large-scale dataset named PENS (PErsonalized News headlineS). The training set is collected from user impressions logs of Microsoft News, and the test set is manually created by hundreds of native speakers to enable a fair testbed for evaluating models in an offline mode. We propose a generic framework as a preparatory solution to our problem. At its heart, user preference is learned by leveraging the user behavioral data, and three kinds of user preference injections are proposed to personalize a text generator and establish personalized headlines. We investigate our dataset by implementing several state-of-the-art user modeling methods in our framework to demonstrate a benchmark score for the proposed dataset. The dataset is available at https://msnews.github.io/pens.html.

Recent pretraining models in Chinese neglect two important aspects specific to the Chinese language: glyph and pinyin, which carry significant syntax and semantic information for language understanding. In this work, we propose ChineseBERT, which incorporates both the glyph and pinyin information of Chinese characters into language model pretraining. The glyph embedding is obtained based on different fonts of a Chinese character, being able to capture character semantics from the visual features, and the pinyin embedding characterizes the pronunciation of Chinese characters, which handles the highly prevalent heteronym phenomenon in Chinese (the same character has different pronunciations with different meanings). Pretrained on large-scale unlabeled Chinese corpus, the proposed ChineseBERT model yields significant performance boost over baseline models with fewer training steps. The proposed model achieves new SOTA performances on a wide range of Chinese NLP tasks, including machine reading comprehension, natural language inference, text classification, sentence pair matching, and competitive performances in named entity recognition and word segmentation.

It has been widely recognized that syntax information can help end-to-end neural machine translation (NMT) systems to achieve better translation. In order to integrate dependency information into Transformer based NMT, existing approaches either exploit words’ local head-dependent relations, ignoring their non-local neighbors carrying important context; or approximate two words’ syntactic relation by their relative distance on the dependency tree, sacrificing exactness. To address these issues, we propose global positional encoding for dependency tree, a new scheme that facilitates syntactic relation modeling between any two words with keeping exactness and without immediate neighbor constraint. Experiment results on NC11 German→English, English→German and WMT English→German datasets show that our approach is more effective than the above two strategies. In addition, our experiments quantitatively show that compared with higher layers, lower layers of the model are more proper places to incorporate syntax information in terms of each layer’s preference to the syntactic pattern and the final performance.

In this work, we re-examine the problem of extractive text summarization for long documents. We observe that the process of extracting summarization of human can be divided into two stages: 1) a rough reading stage to look for sketched information, and 2) a subsequent careful reading stage to select key sentences to form the summary. By simulating such a two-stage process, we propose a novel approach for extractive summarization. We formulate the problem as a contextual-bandit problem and solve it with policy gradient. We adopt a convolutional neural network to encode gist of paragraphs for rough reading, and a decision making policy with an adapted termination mechanism for careful reading. Experiments on the CNN and DailyMail datasets show that our proposed method can provide high-quality summaries with varied length, and significantly outperform the state-of-the-art extractive methods in terms of ROUGE metrics.

The rapid development of knowledge graphs (KGs), such as Freebase and WordNet, has changed the paradigm for AI-related applications. However, even though these KGs are impressively large, most of them are suffering from incompleteness, which leads to performance degradation of AI applications. Most existing researches are focusing on knowledge graph embedding (KGE) models. Nevertheless, those models simply embed entities and relations into latent vectors without leveraging the rich information from the relation structure. Indeed, relations in KGs conform to a three-layer hierarchical relation structure (HRS), i.e., semantically similar relations can make up relation clusters and some relations can be further split into several fine-grained sub-relations. Relation clusters, relations and sub-relations can fit in the top, the middle and the bottom layer of three-layer HRS respectively. To this end, in this paper, we extend existing KGE models TransE, TransH and DistMult, to learn knowledge representations by leveraging the information from the HRS. Particularly, our approach is capable to extend other KGE models. Finally, the experiment results clearly validate the effectiveness of the proposed approach against baselines.