arXiv subjects

Hu Xu

Publications and source records attributed to Hu Xu.

At least 73 records · Page 4Linked to original sources

arXiv2021

CLASSIC: Continual and Contrastive Learning of Aspect Sentiment Classification Tasks

This paper studies continual learning (CL) of a sequence of aspect sentiment classification(ASC) tasks in a particular CL setting called domain incremental learning (DIL). Each task is from a different domain or product. The DIL setting is particularly suited to ASC because in testing the system needs not know the task/domain to which the test data belongs. To our knowledge, this setting has not been studied before for ASC. This paper proposes a novel model called CLASSIC. The key novelty is a contrastive continual learning method that enables both knowledge transfer across tasks and knowledge distillation from old tasks to the new task, which eliminates the need for task ids in testing. Experimental results show the high effectiveness of CLASSIC.

cs.CL↗

arXiv2021

Netmarble AI Center's WMT21 Automatic Post-Editing Shared Task Submission

This paper describes Netmarble's submission to WMT21 Automatic Post-Editing (APE) Shared Task for the English-German language pair. First, we propose a Curriculum Training Strategy in training stages. Facebook Fair's WMT19 news translation model was chosen to engage the large and powerful pre-trained neural networks. Then, we post-train the translation model with different levels of data at each training stages. As the training stages go on, we make the system learn to solve multiple tasks by adding extra information at different training stages gradually. We also show a way to utilize the additional data in large volume for APE tasks. For further improvement, we apply Multi-Task Learning Strategy with the Dynamic Weight Average during the fine-tuning stage. To fine-tune the APE corpus with limited data, we add some related subtasks to learn a unified representation. Finally, for better performance, we leverage external translations as augmented machine translation (MT) during the post-training and fine-tuning. As experimental results show, our APE system significantly improves the translations of provided MT results by -2.848 and +3.74 on the development dataset in terms of TER and BLEU, respectively. It also demonstrates its effectiveness on the test dataset with higher quality than the development dataset.

cs.CL↗

arXiv2021

VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding

We present VideoCLIP, a contrastive approach to pre-train a unified model for zero-shot video and text understanding, without using any labels on downstream tasks. VideoCLIP trains a transformer for video and text by contrasting temporally overlapping positive video-text pairs with hard negatives from nearest neighbor retrieval. Our experiments on a diverse series of downstream tasks, including sequence-level text-video retrieval, VideoQA, token-level action localization, and action segmentation reveal state-of-the-art performance, surpassing prior work, and in some cases even outperforming supervised approaches. Code is made available at https://github.com/pytorch/fairseq/tree/main/examples/MMPT.

cs.CV↗

arXiv2021

VLM: Task-agnostic Video-Language Model Pre-training for Video Understanding

We present a simplified, task-agnostic multi-modal pre-training approach that can accept either video or text input, or both for a variety of end tasks. Existing pre-training are task-specific by adopting either a single cross-modal encoder that requires both modalities, limiting their use for retrieval-style end tasks or more complex multitask learning with two unimodal encoders, limiting early cross-modal fusion. We instead introduce new pretraining masking schemes that better mix across modalities (e.g. by forcing masks for text to predict the closest video embeddings) while also maintaining separability (e.g. unimodal predictions are sometimes required, without using all the input). Experimental results show strong performance across a wider range of tasks than any previous methods, often outperforming task-specific pre-training. Code is made available at https://github.com/pytorch/fairseq/tree/main/examples/MMPT.

cs.CV↗

arXiv2021

NUANCED: Natural Utterance Annotation for Nuanced Conversation with Estimated Distributions

Existing conversational systems are mostly agent-centric, which assumes the user utterances would closely follow the system ontology (for NLU or dialogue state tracking). However, in real-world scenarios, it is highly desirable that the users can speak freely in their own way. It is extremely hard, if not impossible, for the users to adapt to the unknown system ontology. In this work, we attempt to build a user-centric dialogue system. As there is no clean mapping for a user's free form utterance to an ontology, we first model the user preferences as estimated distributions over the system ontology and map the users' utterances to such distributions. Learning such a mapping poses new challenges on reasoning over existing knowledge, ranging from factoid knowledge, commonsense knowledge to the users' own situations. To this end, we build a new dataset named NUANCED that focuses on such realistic settings for conversational recommendation. Collected via dialogue simulation and paraphrasing, NUANCED contains 5.1k dialogues, 26k turns of high-quality user responses. We conduct experiments, showing both the usefulness and challenges of our problem setting. We believe NUANCED can serve as a valuable resource to push existing research from the agent-centric system to the user-centric system. The code and data is publicly available at \url{https://github.com/facebookresearch/nuanced}.

cs.CL↗

arXiv2021

HTLM: Hyper-Text Pre-Training and Prompting of Language Models

We introduce HTLM, a hyper-text language model trained on a large-scale web crawl. Modeling hyper-text has a number of advantages: (1) it is easily gathered at scale, (2) it provides rich document-level and end-task-adjacent supervision (e.g. class and id attributes often encode document category information), and (3) it allows for new structured prompting that follows the established semantics of HTML (e.g. to do zero-shot summarization by infilling title tags for a webpage that contains the input text). We show that pretraining with a BART-style denoising loss directly on simplified HTML provides highly effective transfer for a wide range of end tasks and supervision levels. HTLM matches or exceeds the performance of comparably sized text-only LMs for zero-shot prompting and fine-tuning for classification benchmarks, while also setting new state-of-the-art performance levels for zero-shot summarization. We also find that hyper-text prompts provide more value to HTLM, in terms of data efficiency, than plain text prompts do for existing LMs, and that HTLM is highly effective at auto-prompting itself, by simply generating the most likely hyper-text formatting for any available training data. We will release all code and models to support future HTLM research.

cs.CL↗

arXiv2021

Three-dimensional Dirac Phonons with Inversion Symmetry

Dirac semimetals associated with bulk Dirac fermions are well-known in topological electronic systems. In sharp contrast, three-dimensional (3D) Dirac phonons in crystalline solids are still unavailable. Here we perform symmetry arguments and first-principles calculations to systematically investigate 3D Dirac phonons in all space groups with inversion symmetry. The results show that there are two categories of 3D Dirac phonons depending on their protection mechanisms and positions in momentum space. The first category originates from the four-dimensional irreducible representations at the high symmetry points. The second category arises from the phonon branch inversion, and the symmetry guarantees Dirac points to be located along the high symmetry lines. Furthermore, we reveal that non-symmorphic symmetries and the combination of inversion and time-reversal symmetries play essential roles in the emergence of 3D Dirac phonons. Our work not only offers a comprehensive understanding of 3D Dirac phonons but also provides significant guidance for exploring Dirac bosons in both phononic and photonic systems.

cond-mat.mtrl-sci↗

arXiv2021

Straight nodal-line phonons in symmorphic space groups

Based on first-principles calculations and effective model analysis, all possible straight nodal-line phonons in symmorphic space groups are uncovered. A classification of two-band ${\textbf{k}}\cdot {\textbf{p}}$ models at an arbitrary point on the nodal lines is carried out, and these nodal-line phonons are found to locate along the high symmetry lines of the Brillouin zone. According to our classification, there are two types of nodal-line phonons in symmorphic space groups. The first type possesses a linear dispersion perpendicular to the nodal lines, while the second type exhibits a quadratic dispersion. The underlying mechanisms are revealed by symmetry analysis, and the types of nodal-line phonons can be distinguished through the little group of the high symmetry lines. Our work not only classifies nodal-line phonons but also provides guidance to search for topological nodal lines in bosonic systems.

cond-mat.mtrl-sci↗

arXiv2020

Understanding Pre-trained BERT for Aspect-based Sentiment Analysis

This paper analyzes the pre-trained hidden representations learned from reviews on BERT for tasks in aspect-based sentiment analysis (ABSA). Our work is motivated by the recent progress in BERT-based language models for ABSA. However, it is not clear how the general proxy task of (masked) language model trained on unlabeled corpus without annotations of aspects or opinions can provide important features for downstream tasks in ABSA. By leveraging the annotated datasets in ABSA, we investigate both the attentions and the learned representations of BERT pre-trained on reviews. We found that BERT uses very few self-attention heads to encode context words (such as prepositions or pronouns that indicating an aspect) and opinion words for an aspect. Most features in the representation of an aspect are dedicated to the fine-grained semantics of the domain (or product category) and the aspect itself, instead of carrying summarized opinions from its context. We hope this investigation can help future research in improving self-supervised learning, unsupervised learning and fine-tuning for ABSA. The pre-trained model and code can be found at https://github.com/howardhsu/BERT-for-RRC-ABSA.

cs.CL↗

arXiv2020

Controllable Text Generation with Focused Variation

This work introduces Focused-Variation Network (FVN), a novel model to control language generation. The main problems in previous controlled language generation models range from the difficulty of generating text according to the given attributes, to the lack of diversity of the generated texts. FVN addresses these issues by learning disjoint discrete latent spaces for each attribute inside codebooks, which allows for both controllability and diversity, while at the same time generating fluent text. We evaluate FVN on two text generation datasets with annotated content and style, and show state-of-the-art performance as assessed by automatic and human evaluations.

cs.CL↗

arXiv2020

User Memory Reasoning for Conversational Recommendation

We study a conversational recommendation model which dynamically manages users' past (offline) preferences and current (online) requests through a structured and cumulative user memory knowledge graph, to allow for natural interactions and accurate recommendations. For this study, we create a new Memory Graph (MG) <--> Conversational Recommendation parallel corpus called MGConvRex with 7K+ human-to-human role-playing dialogs, grounded on a large-scale user memory bootstrapped from real-world user scenarios. MGConvRex captures human-level reasoning over user memory and has disjoint training/testing sets of users for zero-shot (cold-start) reasoning for recommendation. We propose a simple yet expandable formulation for constructing and updating the MG, and a reasoning model that predicts optimal dialog policies and recommendation items in unconstrained graph space. The prediction of our proposed model inherits the graph structure, providing a natural way to explain the model's recommendation. Experiments are conducted for both offline metrics and online simulation, showing competitive results.

cs.CL↗

arXiv2020

DomBERT: Domain-oriented Language Model for Aspect-based Sentiment Analysis

This paper focuses on learning domain-oriented language models driven by end tasks, which aims to combine the worlds of both general-purpose language models (such as ELMo and BERT) and domain-specific language understanding. We propose DomBERT, an extension of BERT to learn from both in-domain corpus and relevant domain corpora. This helps in learning domain language models with low-resources. Experiments are conducted on an assortment of tasks in aspect-based sentiment analysis, demonstrating promising results.

cs.CL↗

arXiv2019

Hybridization-Induced Gapped and Gapless States on the Surfaces of Magnetic Topological Insulators

The layered MnBi2nTe3n+1 family represents the first intrinsic antiferromagnetic topological insulator (AFM TI, protected by a combination symmetry ) ever discovered, providing an ideal platform to explore novel physics such as quantum anomalous Hall effect at elevated temperature and axion electrodynamics. Recent angle-resolved photoemission spectroscopy (ARPES) experiments on this family have revealed that all terminations exhibit (nearly) gapless topological surface states (TSSs) within the AFM state, violating the definition of the AFM TI, as the surfaces being studied should be -breaking and opening a gap. Here we explain this curious paradox using a surface-bulk band hybridization picture. Combining ARPES and first-principles calculations, we prove that only an apparent gap is opened by hybridization between TSSs and bulk bands. The observed (nearly) gapless features are consistently reproduced by tight-binding simulations where TSSs are coupled to a Rashba-split bulk band. The Dirac-cone-like spectral features are actually of bulk origin, thus not sensitive to the-breaking at the AFM surfaces. This picture explains the (nearly) gapless behaviour found in both Bi2Te3- and MnBi2Te4-terminated surfaces and is applicable to all terminations of MnBi2nTe3n+1 family. Our findings highlight the role of band hybridization, superior to magnetism in this case, in shaping the general surface band structure in magnetic topological materials for the first time.

cond-mat.str-el↗

arXiv2019

Review Conversational Reading Comprehension

Inspired by conversational reading comprehension (CRC), this paper studies a novel task of leveraging reviews as a source to build an agent that can answer multi-turn questions from potential consumers of online businesses. We first build a review CRC dataset and then propose a novel task-aware pre-tuning step running between language model (e.g., BERT) pre-training and domain-specific fine-tuning. The proposed pre-tuning requires no data annotation, but can greatly enhance the performance on our end task. Experimental results show that the proposed approach is highly effective and has competitive performance as the supervised approach. The dataset is available at \url{https://github.com/howardhsu/RCRC}

cs.CL↗

arXiv2019

A Failure of Aspect Sentiment Classifiers and an Adaptive Re-weighting Solution

Aspect-based sentiment classification (ASC) is an important task in fine-grained sentiment analysis.~Deep supervised ASC approaches typically model this task as a pair-wise classification task that takes an aspect and a sentence containing the aspect and outputs the polarity of the aspect in that sentence. However, we discovered that many existing approaches fail to learn an effective ASC classifier but more like a sentence-level sentiment classifier because they have difficulty to handle sentences with different polarities for different aspects.~This paper first demonstrates this problem using several state-of-the-art ASC models. It then proposes a novel and general adaptive re-weighting (ARW) scheme to adjust the training to dramatically improve ASC for such complex sentences. Experimental results show that the proposed framework is effective \footnote{The dataset and code are available at \url{https://github.com/howardhsu/ASC_failure}.}.

cs.CL↗

arXiv2019

Modeling Multi-Action Policy for Task-Oriented Dialogues

Dialogue management (DM) plays a key role in the quality of the interaction with the user in a task-oriented dialogue system. In most existing approaches, the agent predicts only one DM policy action per turn. This significantly limits the expressive power of the conversational agent and introduces unwanted turns of interactions that may challenge users' patience. Longer conversations also lead to more errors and the system needs to be more robust to handle them. In this paper, we compare the performance of several models on the task of predicting multiple acts for each turn. A novel policy model is proposed based on a recurrent cell called gated Continue-Act-Slots (gCAS) that overcomes the limitations of the existing models. Experimental results show that gCAS outperforms other approaches. The code is available at https://leishu02.github.io/

cs.CL↗

arXiv2019

Quantum Confined Tomonaga-Luttinger Liquid in MoSe2 Twin Domain Boundaries

There have been conflicting reports on the electronic properties of twin domain boundaries (DBs) in MoSe2 monolayer, including the quantum well states, charge density wave, and Tomonaga-Luttinger liquid (TLL). Here we employ low-temperature scanning tunneling spectroscopy to reveal both the quantum confinement effect and signatures of TLL in the one-dimensional DBs. The data do not support the CDW at temperatures down to ~5 K.

cond-mat.mes-hall↗

arXiv2019

Flexibly-Structured Model for Task-Oriented Dialogues

This paper proposes a novel end-to-end architecture for task-oriented dialogue systems. It is based on a simple and practical yet very effective sequence-to-sequence approach, where language understanding and state tracking tasks are modeled jointly with a structured copy-augmented sequential decoder and a multi-label decoder for each slot. The policy engine and language generation tasks are modeled jointly following that. The copy-augmented sequential decoder deals with new or unknown values in the conversation, while the multi-label decoder combined with the sequential decoder ensures the explicit assignment of values to slots. On the generation part, slot binary classifiers are used to improve performance. This architecture is scalable to real-world scenarios and is shown through an empirical evaluation to achieve state-of-the-art performance on both the Cambridge Restaurant dataset and the Stanford in-car assistant dataset\footnote{The code is available at \url{https://github.com/uber-research/FSDM}}

cs.CL↗