arXiv subjects

Yang Feng

Publications and source records attributed to Yang Feng.

At least 199 records · Page 11Linked to original sources

arXiv2021

Towards Expressive Communication with Internet Memes: A New Multimodal Conversation Dataset and Benchmark

As a kind of new expression elements, Internet memes are popular and extensively used in online chatting scenarios since they manage to make dialogues vivid, moving, and interesting. However, most current dialogue researches focus on text-only dialogue tasks. In this paper, we propose a new task named as \textbf{M}eme incorporated \textbf{O}pen-domain \textbf{D}ialogue (MOD). Compared to previous dialogue tasks, MOD is much more challenging since it requires the model to understand the multimodal elements as well as the emotions behind them. To facilitate the MOD research, we construct a large-scale open-domain multimodal dialogue dataset incorporating abundant Internet memes into utterances. The dataset consists of $\sim$45K Chinese conversations with $\sim$606K utterances. Each conversation contains about $13$ utterances with about $4$ Internet memes on average and each utterance equipped with an Internet meme is annotated with the corresponding emotion. In addition, we present a simple and effective method, which utilizes a unified generation network to solve the MOD task. Experimental results demonstrate that our method trained on the proposed corpus is able to achieve expressive communication including texts and memes. The corpus and models have been publicly available at https://github.com/lizekang/DSTC10-MOD.

cs.CL↗

arXiv2021

Sequence-Level Training for Non-Autoregressive Neural Machine Translation

In recent years, Neural Machine Translation (NMT) has achieved notable results in various translation tasks. However, the word-by-word generation manner determined by the autoregressive mechanism leads to high translation latency of the NMT and restricts its low-latency applications. Non-Autoregressive Neural Machine Translation (NAT) removes the autoregressive mechanism and achieves significant decoding speedup through generating target words independently and simultaneously. Nevertheless, NAT still takes the word-level cross-entropy loss as the training objective, which is not optimal because the output of NAT cannot be properly evaluated due to the multimodality problem. In this article, we propose using sequence-level training objectives to train NAT models, which evaluate the NAT outputs as a whole and correlates well with the real translation quality. Firstly, we propose training NAT models to optimize sequence-level evaluation metrics (e.g., BLEU) based on several novel reinforcement algorithms customized for NAT, which outperforms the conventional method by reducing the variance of gradient estimation. Secondly, we introduce a novel training objective for NAT models, which aims to minimize the Bag-of-Ngrams (BoN) difference between the model output and the reference sentence. The BoN training objective is differentiable and can be calculated efficiently without doing any approximations. Finally, we apply a three-stage training strategy to combine these two methods to train the NAT model. We validate our approach on four translation tasks (WMT14 En$\leftrightarrow$De, WMT16 En$\leftrightarrow$Ro), which shows that our approach largely outperforms NAT baselines and achieves remarkable performance on all translation tasks. The source code is available at https://github.com/ictnlp/Seq-NAT.

cs.CL↗

arXiv2021

Importance-based Neuron Allocation for Multilingual Neural Machine Translation

Multilingual neural machine translation with a single model has drawn much attention due to its capability to deal with multiple languages. However, the current multilingual translation paradigm often makes the model tend to preserve the general knowledge, but ignore the language-specific knowledge. Some previous works try to solve this problem by adding various kinds of language-specific modules to the model, but they suffer from the parameter explosion problem and require specialized manual design. To solve these problems, we propose to divide the model neurons into general and language-specific parts based on their importance across languages. The general part is responsible for preserving the general knowledge and participating in the translation of all the languages, while the language-specific part is responsible for preserving the language-specific knowledge and participating in the translation of some specific languages. Experimental results on several language pairs, covering IWSLT and Europarl corpus datasets, demonstrate the effectiveness and universality of the proposed method.

cs.CL↗

arXiv2021

Imbalanced classification: a paradigm-based review

A common issue for classification in scientific research and industry is the existence of imbalanced classes. When sample sizes of different classes are imbalanced in training data, naively implementing a classification method often leads to unsatisfactory prediction results on test data. Multiple resampling techniques have been proposed to address the class imbalance issues. Yet, there is no general guidance on when to use each technique. In this article, we provide a paradigm-based review of the common resampling techniques for binary classification under imbalanced class sizes. The paradigms we consider include the classical paradigm that minimizes the overall classification error, the cost-sensitive learning paradigm that minimizes a cost-adjusted weighted type I and type II errors, and the Neyman-Pearson paradigm that minimizes the type II error subject to a type I error constraint. Under each paradigm, we investigate the combination of the resampling techniques and a few state-of-the-art classification methods. For each pair of resampling techniques and classification methods, we use simulation studies and a real data set on credit card fraud to study the performance under different evaluation metrics. From these extensive numerical experiments, we demonstrate under each classification paradigm, the complex dynamics among resampling techniques, base classification methods, evaluation metrics, and imbalance ratios. We also summarize a few takeaway messages regarding the choices of resampling techniques and base classification methods, which could be helpful for practitioners.

stat.ME↗

arXiv2021

Guiding Teacher Forcing with Seer Forcing for Neural Machine Translation

Although teacher forcing has become the main training paradigm for neural machine translation, it usually makes predictions only conditioned on past information, and hence lacks global planning for the future. To address this problem, we introduce another decoder, called seer decoder, into the encoder-decoder framework during training, which involves future information in target predictions. Meanwhile, we force the conventional decoder to simulate the behaviors of the seer decoder via knowledge distillation. In this way, at test the conventional decoder can perform like the seer decoder without the attendance of it. Experiment results on the Chinese-English, English-German and English-Romanian translation tasks show our method can outperform competitive baselines significantly and achieves greater improvements on the bigger data sets. Besides, the experiments also prove knowledge distillation the best way to transfer knowledge from the seer decoder to the conventional decoder compared to adversarial learning and L2 regularization.

cs.CL↗

arXiv2021

GTM: A Generative Triple-Wise Model for Conversational Question Generation

Generating some appealing questions in open-domain conversations is an effective way to improve human-machine interactions and lead the topic to a broader or deeper direction. To avoid dull or deviated questions, some researchers tried to utilize answer, the "future" information, to guide question generation. However, they separate a post-question-answer (PQA) triple into two parts: post-question (PQ) and question-answer (QA) pairs, which may hurt the overall coherence. Besides, the QA relationship is modeled as a one-to-one mapping that is not reasonable in open-domain conversations. To tackle these problems, we propose a generative triple-wise model with hierarchical variations for open-domain conversational question generation (CQG). Latent variables in three hierarchies are used to represent the shared background of a triple and one-to-many semantic mappings in both PQ and QA pairs. Experimental results on a large-scale CQG dataset show that our method significantly improves the quality of questions in terms of fluency, coherence and diversity over competitive baselines.

cs.CL↗

arXiv2021

Testing Community Structures for Hypergraphs

Many complex networks in real world can be formulated as hypergraphs where community detection has been widely used. However, the fundamental question of whether communities exist or not in an observed hypergraph still remains unresolved. The aim of the present paper is to tackle this important problem. Specifically, we study when a hypergraph with community structure can be successfully distinguished from its Erdös-Renyi counterpart, and propose concrete test statistics based on hypergraph cycles when the models are distinguishable. Our contributions are summarized as follows. For uniform hypergraphs, we show that successful testing is always impossible when average degree tends to zero, might be possible when average degree is bounded, and is possible when average degree is growing. We obtain asymptotic distributions of the proposed test statistics and analyze their power. Our results for growing degree case are further extended to nonuniform hypergraphs in which a new test involving both edge and hyperedge information is proposed. The novel aspect of our new test is that it is provably more powerful than the classic test involving only edge information. Simulation and real data analysis support our theoretical findings. The proofs rely on Janson's contiguity theory (\cite{J95}) and a high-moments driven asymptotic normality result by Gao and Wormald (\cite{GWALD}).

math.ST↗

arXiv2021

WeChat AI & ICT's Submission for DSTC9 Interactive Dialogue Evaluation Track

We participate in the DSTC9 Interactive Dialogue Evaluation Track (Gunasekara et al. 2020) sub-task 1 (Knowledge Grounded Dialogue) and sub-task 2 (Interactive Dialogue). In sub-task 1, we employ a pre-trained language model to generate topic-related responses and propose a response ensemble method for response selection. In sub-task2, we propose a novel Dialogue Planning Model (DPM) to capture conversation flow in the interaction with humans. We also design an integrated open-domain dialogue system containing pre-process, dialogue model, scoring model, and post-process, which can generate fluent, coherent, consistent, and humanlike responses. We tie 1st on human ratings and also get the highest Meteor, and Bert-score in sub-task 1, and rank 3rd on interactive human evaluation in sub-task 2.

cs.CL↗

arXiv2021

Conversations Are Not Flat: Modeling the Dynamic Information Flow across Dialogue Utterances

Nowadays, open-domain dialogue models can generate acceptable responses according to the historical context based on the large-scale pre-trained language models. However, they generally concatenate the dialogue history directly as the model input to predict the response, which we named as the flat pattern and ignores the dynamic information flow across dialogue utterances. In this work, we propose the DialoFlow model, in which we introduce a dynamic flow mechanism to model the context flow, and design three training objectives to capture the information dynamics across dialogue utterances by addressing the semantic influence brought about by each utterance in large-scale pre-training. Experiments on the multi-reference Reddit Dataset and DailyDialog Dataset demonstrate that our DialoFlow significantly outperforms the DialoGPT on the dialogue generation task. Besides, we propose the Flow score, an effective automatic metric for evaluating interactive human-bot conversation quality based on the pre-trained DialoFlow, which presents high chatbot-level correlation ($r=0.9$) with human ratings among 11 chatbots. Code and pre-trained models will be public. \footnote{\url{https://github.com/ictnlp/DialoFlow}}

cs.CL↗

arXiv2021

Addressing Inquiries about History: An Efficient and Practical Framework for Evaluating Open-domain Chatbot Consistency

A good open-domain chatbot should avoid presenting contradictory responses about facts or opinions in a conversational session, known as its consistency capacity. However, evaluating the consistency capacity of a chatbot is still challenging. Employing human judges to interact with chatbots on purpose to check their capacities is costly and low-efficient, and difficult to get rid of subjective bias. In this paper, we propose the Addressing Inquiries about History (AIH), an efficient and practical framework for the consistency evaluation. At the conversation stage, AIH attempts to address appropriate inquiries about the dialogue history to induce the chatbot to redeclare the historical facts or opinions. We carry out the conversation between chatbots, which is more efficient than the human-bot interaction and can also alleviate the subjective bias. In this way, we manage to rapidly obtain a dialog session that contains responses with high contradiction possibilities. At the contradiction recognition stage, we can either employ human judges or a natural language inference (NLI) model to recognize whether the answers to the inquiries are contradictory with history. Finally, we are able to rank chatbots according to the contradiction statistics. Experiments on open-domain chatbots show that our approach can efficiently and reliably assess the consistency capacity of chatbots and achieve a high ranking correlation with the human evaluation. We release the framework and hope to help improve the consistency capacity of chatbots. \footnote{\url{https://github.com/ictnlp/AIH}}

cs.CL↗

arXiv2021

RaSE: Random Subspace Ensemble Classification

We propose a flexible ensemble classification framework, Random Subspace Ensemble (RaSE), for sparse classification. In the RaSE algorithm, we aggregate many weak learners, where each weak learner is a base classifier trained in a subspace optimally selected from a collection of random subspaces. To conduct subspace selection, we propose a new criterion, ratio information criterion (RIC), based on weighted Kullback-Leibler divergence. The theoretical analysis includes the risk and Monte-Carlo variance of the RaSE classifier, establishing the screening consistency and weak consistency of RIC, and providing an upper bound for the misclassification rate of the RaSE classifier. In addition, we show that in a high-dimensional framework, the number of random subspaces needs to be very large to guarantee that a subspace covering signals is selected. Therefore, we propose an iterative version of the RaSE algorithm and prove that under some specific conditions, a smaller number of generated random subspaces are needed to find a desirable subspace through iteration. An array of simulations under various models and real-data applications demonstrate the effectiveness and robustness of the RaSE classifier and its iterative version in terms of low misclassification rate and accurate feature ranking. The RaSE algorithm is implemented in the R package RaSEn on CRAN.

stat.ML↗

arXiv2021

RaSE: A Variable Screening Framework via Random Subspace Ensembles

Variable screening methods have been shown to be effective in dimension reduction under the ultra-high dimensional setting. Most existing screening methods are designed to rank the predictors according to their individual contributions to the response. As a result, variables that are marginally independent but jointly dependent with the response could be missed. In this work, we propose a new framework for variable screening, Random Subspace Ensemble (RaSE), which works by evaluating the quality of random subspaces that may cover multiple predictors. This new screening framework can be naturally combined with any subspace evaluation criterion, which leads to an array of screening methods. The framework is capable to identify signals with no marginal effect or with high-order interaction effects. It is shown to enjoy the sure screening property and rank consistency. We also develop an iterative version of RaSE screening with theoretical support. Extensive simulation studies and real-data analysis show the effectiveness of the new screening framework.

stat.ME↗

arXiv2021

SE-DAE: Style-Enhanced Denoising Auto-Encoder for Unsupervised Text Style Transfer

Text style transfer aims to change the style of sentences while preserving the semantic meanings. Due to the lack of parallel data, the Denoising Auto-Encoder (DAE) is widely used in this task to model distributions of different sentence styles. However, because of the conflict between the target of the conventional denoising procedure and the target of style transfer task, the vanilla DAE can not produce satisfying enough results. To improve the transferability of the model, most of the existing works combine DAE with various complicated unsupervised networks, which makes the whole system become over-complex. In this work, we design a novel DAE model named Style-Enhanced DAE (SE-DAE), which is specifically designed for the text style transfer task. Compared with previous complicated style-transfer models, our model do not consist of any complicated unsupervised networks, but only relies on the high-quality pseudo-parallel data generated by a novel data refinement mechanism. Moreover, to alleviate the conflict between the targets of the conventional denoising procedure and the style transfer task, we propose another novel style denoising mechanism, which is more compatible with the target of the style transfer task. We validate the effectiveness of our model on two style benchmark datasets. Both automatic evaluation and human evaluation show that our proposed model is highly competitive compared with previous strong the state of the art (SOTA) approaches and greatly outperforms the vanilla DAE.

cs.CL↗

arXiv2021

Modeling Coverage for Non-Autoregressive Neural Machine Translation

Non-Autoregressive Neural Machine Translation (NAT) has achieved significant inference speedup by generating all tokens simultaneously. Despite its high efficiency, NAT usually suffers from two kinds of translation errors: over-translation (e.g. repeated tokens) and under-translation (e.g. missing translations), which eventually limits the translation quality. In this paper, we argue that these issues of NAT can be addressed through coverage modeling, which has been proved to be useful in autoregressive decoding. We propose a novel Coverage-NAT to model the coverage information directly by a token-level coverage iterative refinement mechanism and a sentence-level coverage agreement, which can remind the model if a source token has been translated or not and improve the semantics consistency between the translation and the source, respectively. Experimental results on WMT14 En-De and WMT16 En-Ro translation tasks show that our method can alleviate those errors and achieve strong improvements over the baseline system.

cs.CL↗

arXiv2021

Pruning-then-Expanding Model for Domain Adaptation of Neural Machine Translation

Domain Adaptation is widely used in practical applications of neural machine translation, which aims to achieve good performance on both the general-domain and in-domain. However, the existing methods for domain adaptation usually suffer from catastrophic forgetting, domain divergence, and model explosion. To address these three problems, we propose a method of "divide and conquer" which is based on the importance of neurons or parameters in the translation model. In our method, we first prune the model and only keep the important neurons or parameters, making them responsible for both general-domain and in-domain translation. Then we further train the pruned model supervised by the original unpruned model with the knowledge distillation method. Last we expand the model to the original size and fine-tune the added parameters for the in-domain translation. We conduct experiments on different languages and domains and the results show that our method can achieve significant improvements compared with several strong baselines.

cs.CL↗

arXiv2021

Cross-status Communication and Project Outcomesin OSS Development A Language Style Matching Perspective

The success of an open source software (OSS) project requires effective communication among its members. Given that OSS projects often have established social status systems, much communication may happen between individuals of different statuses, particularly, elite developers who have project management privileges and ordinary project contributors. Since sociolinguistics literature and our prior work have found that groups in different status would be likely develop different language styles, which may hinder critical cross status communication, and hereby influence the outcomes of the project. We seek to develop an understanding of cross status communication, as well as its impacts on the outcomes of an OSS project in terms of productivity and quality. The anticipated results describe the linguistic similarities and differences of cross status communication and reveal the relationships between these linguistic similarities and differences and project outcomes.

cs.SE↗

arXiv2021

Distinct quantum anomalous Hall ground states induced by magnetic disorders

The quantum anomalous Hall (QAH) effect in magnetic topological insulator (TI) represents a new state of matter originated from the interplay between topology and magnetism. The defining characteristics of the QAH ground state are the quantized Hall resistivity ($ρ_{yx}$) and vanishing longitudinal resistivity ($ρ_{xx}$) in the absence of external magnetic field. A fundamental question concerning the QAH effect is whether it is merely a zero-magnetic-field quantum Hall (QH) effect, or if it can host unique quantum phases and phase transitions that are unavailable elsewhere. The most dramatic departure of the QAH systems from other QH systems lies in the strong magnetic disorders that induce spatially random magnetization. Because disorder and magnetism play pivotal roles in the phase diagram of two-dimensional electron systems, the high degree of magnetic disorders in QAH systems may create novel phases and quantum critical phenomena. In this work, we perform systematic transport studies of a series of magnetic TIs with varied strength of magnetic disorders. We find that the ground state of QAH effect can be categorized into two distinct classes: the QAH insulator and anomalous Hall (AH) insulator phases, as the zero-magnetic-field counterparts of the QH liquid and Hall insulator in the QH systems. In the low disorder limit of the QAH insulator regime, we observe a universal quantized longitudinal resistance $ρ_{xx} = h/e^{2}$ at the coercive field. In the AH insulator regime, we find that a magnetic field can drive it to the QAH insulator phase through a quantum critical point with distinct scaling behaviors from that in the QH phase transition. We propose that the transmission between chiral edge states at domain boundaries, tunable by disorder and magnetic fields, is the key for determining the QAH ground state.

cond-mat.mes-hall↗

arXiv2021

Gate Tunable Supercurrent in Josephson Junctions Based on Bi2Te3 Topological Insulator Thin Films

We report transport measurements on Josephson junctions consisting of Bi2Te3 topological insulator (TI) thin films contacted by superconducting Nb electrodes. For a device with junction length L = 134 nm, the critical supercurrent Ic can be modulated by an electrical gate which tunes the carrier type and density of the TI film. Ic can reach a minimum when the TI is near the charge neutrality regime with the Fermi energy lying close to the Dirac point of the surface state. In the p-type regime the Josephson current can be well described by a short ballistic junction model. In the n-type regime the junction is ballistic at 0.7 K < T < 3.8 K while for T < 0.7 K the diffusive bulk modes emerge and contribute a larger Ic than the ballistic model. We attribute the lack of diffusive bulk modes in the p-type regime to the formation of p-n junctions. Our work provides new clues for search of Majorana zero mode in TI-based superconducting devices.

cond-mat.supr-con↗