Jun Wang


2024

pdf bib
Exploring Union and Intersection of Visual Regions for Generating Questions, Answers, and Distractors
Wenjian Ding | Yao Zhang | Jun Wang | Adam Jatowt | Zhenglu Yang
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing

Multiple-choice visual question answering (VQA) is to automatically choose a correct answer from a set of choices after reading an image. Existing efforts have been devoted to a separate generation of an image-related question, a correct answer, or challenge distractors. By contrast, we turn to a holistic generation and optimization of questions, answers, and distractors (QADs) in this study. This integrated generation strategy eliminates the need for human curation and guarantees information consistency. Furthermore, we first propose to put the spotlight on different image regions to diversify QADs. Accordingly, a novel framework ReBo is formulated in this paper. ReBo cyclically generates each QAD based on a recurrent multimodal encoder, and each generation is focusing on a different area of the image compared to those already concerned by the previously generated QADs. In addition to traditional VQA comparisons with state-of-the-art approaches, we also validate the capability of ReBo in generating augmented data to benefit VQA models.

pdf bib
SEEP: Training Dynamics Grounds Latent Representation Search for Mitigating Backdoor Poisoning Attacks
Xuanli He | Qiongkai Xu | Jun Wang | Benjamin I. P. Rubinstein | Trevor Cohn
Transactions of the Association for Computational Linguistics, Volume 12

Modern NLP models are often trained on public datasets drawn from diverse sources, rendering them vulnerable to data poisoning attacks. These attacks can manipulate the model’s behavior in ways engineered by the attacker. One such tactic involves the implantation of backdoors, achieved by poisoning specific training instances with a textual trigger and a target class label. Several strategies have been proposed to mitigate the risks associated with backdoor attacks by identifying and removing suspected poisoned examples. However, we observe that these strategies fail to offer effective protection against several advanced backdoor attacks. To remedy this deficiency, we propose a novel defensive mechanism that first exploits training dynamics to identify poisoned samples with high precision, followed by a label propagation step to improve recall and thus remove the majority of poisoned instances. Compared with recent advanced defense methods, our method considerably reduces the success rates of several backdoor attacks while maintaining high classification accuracy on clean test sets.

pdf bib
Multiple Knowledge-Enhanced Interactive Graph Network for Multimodal Conversational Emotion Recognition
Geng Tu | Jun Wang | Zhenyu Li | Shiwei Chen | Bin Liang | Xi Zeng | Min Yang | Ruifeng Xu
Findings of the Association for Computational Linguistics: EMNLP 2024

Multimodal Emotion Recognition in Conversations (ERC) aims to identify emotions in conversational videos. Current efforts focus on modeling both context-sensitive and speaker-sensitive dependencies and multimodal fusion. Despite the progress, models in Multimodal ERC (MERC) still struggle due to a lack of CommonSense Knowledge (CSK). In contrast, models in textual ERC typically employ CSK to enhance emotion inference. However, in multimodal scenarios, relying solely on textual CSK while neglecting visual CSK may hinder the understanding of visual emotional cues. To address this, we introduce a novel approach called Multiple Knowledge Enhanced Interactive Graph Network (MKE-IGN) to integrate multiple knowledge, such as textual and visual CSK, into the edge representations, thereby facilitating the modeling of relations between utterances and different types of CSK. Furthermore, considering that irrelevant CSK might be retained as noise, MKE-IGN adaptively selects this CSK guided by the mood-congruent effect and refines it based on contexts. Experimental results show that MKE-IGN outperforms state-of-the-art methods on two popular datasets.

pdf bib
Backdoor Attacks on Multilingual Machine Translation
Jun Wang | Qiongkai Xu | Xuanli He | Benjamin Rubinstein | Trevor Cohn
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)

While multilingual machine translation (MNMT) systems hold substantial promise, they also have security vulnerabilities. Our research highlights that MNMT systems can be susceptible to a particularly devious style of backdoor attack, whereby an attacker injects poisoned data into a low-resource language pair to cause malicious translations in other languages, including high-resource languages.Our experimental results reveal that injecting less than 0.01% poisoned data into a low-resource language pair can achieve an average 20% attack success rate in attacking high-resource language pairs. This type of attack is of particular concern, given the larger attack surface of languages inherent to low-resource settings. Our aim is to bring attention to these vulnerabilities within MNMT systems with the hope of encouraging the community to address security concerns in machine translation, especially in the context of low-resource languages.

pdf bib
HITSZ-HLT at WASSA-2024 Shared Task 2: Language-agnostic Multi-task Learning for Explainability of Cross-lingual Emotion Detection
Feng Xiong | Jun Wang | Geng Tu | Ruifeng Xu
Proceedings of the 14th Workshop on Computational Approaches to Subjectivity, Sentiment, & Social Media Analysis

This paper describes the system developed by the HITSZ-HLT team for WASSA-2024 Shared Task 2, which addresses two closely linked sub-tasks: Cross-lingual Emotion Detection and Binary Trigger Word Detection in tweets. The main goal of Shared Task 2 is to simultaneously identify the emotions expressed and detect the trigger words across multiple languages. To achieve this, we introduce a Language-agnostic Multi Task Learning (LaMTL) framework that integrates emotion prediction and emotion trigger word detection tasks. By fostering synergistic interactions between task-specific and task-agnostic representations, the LaMTL aims to mutually enhance emotional cues, ultimately improving the performance of both tasks. Additionally, we leverage large-scale language models to translate the training dataset into multiple languages, thereby fostering the formation of language-agnostic representations within the model, significantly enhancing the model’s ability to transfer and perform well across multilingual data. Experimental results demonstrate the effectiveness of our framework across both tasks, with a particular highlight on its success in achieving second place in sub-task 2.

pdf bib
PocketLLM: Enabling On-Device Fine-Tuning for Personalized LLMs
Dan Peng | Zhihui Fu | Jun Wang
Proceedings of the Fifth Workshop on Privacy in Natural Language Processing

Recent advancements in large language models (LLMs) have indeed showcased their impressive capabilities. On mobile devices, the wealth of valuable, non-public data generated daily holds great promise for locally fine-tuning personalized LLMs, while maintaining privacy through on-device processing. However, the constraints of mobile device resources pose challenges to direct on-device LLM fine-tuning, mainly due to the memory-intensive nature of derivative-based optimization required for saving gradients and optimizer states. To tackle this, we propose employing derivative-free optimization techniques to enable on-device fine-tuning of LLM, even on memory-limited mobile devices. Empirical results demonstrate that the RoBERTa-large model and OPT-1.3B can be fine-tuned locally on the OPPO Reno 6 smartphone using around 4GB and 6.5GB of memory respectively, using derivative-free optimization techniques. This highlights the feasibility of on-device LLM fine-tuning on mobile devices, paving the way for personalized LLMs on resource-constrained devices while safeguarding data privacy.

pdf bib
Self-Contrast: Better Reflection Through Inconsistent Solving Perspectives
Wenqi Zhang | Yongliang Shen | Linjuan Wu | Qiuying Peng | Jun Wang | Yueting Zhuang | Weiming Lu
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

The reflection capacity of Large Language Model (LLM) has garnered extensive attention. A post-hoc prompting strategy, e.g., reflexion and self-refine, refines LLM’s response based on self-evaluated or external feedback. However, recent research indicates without external feedback, LLM’s intrinsic reflection is unstable. Our investigation unveils that the key bottleneck is the quality of the self-evaluated feedback. We find LLMs often exhibit overconfidence or high randomness when self-evaluate, offering stubborn or inconsistent feedback, which causes poor reflection. To remedy this, we advocate Self-Contrast: It adaptively explores diverse solving perspectives tailored to the request, contrasts the differences, and summarizes these discrepancies into a checklist which could be used to re-examine and eliminate discrepancies. Our method endows LLM with diverse perspectives to alleviate stubborn biases. Moreover, their discrepancies indicate potential errors or inherent uncertainties that LLM often overlooks. Reflecting upon these can catalyze more accurate and stable reflection. Experiments conducted on a series of reasoning and translation tasks with different LLMs serve to underscore the effectiveness and generality of our strategy.

pdf bib
D2LLM: Decomposed and Distilled Large Language Models for Semantic Search
Zihan Liao | Hang Yu | Jianguo Li | Jun Wang | Wei Zhang
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

The key challenge in semantic search is to create models that are both accurate and efficient in pinpointing relevant sentences for queries. While BERT-style bi-encoders excel in efficiency with pre-computed embeddings, they often miss subtle nuances in search tasks. Conversely, GPT-style LLMs with cross-encoder designs capture these nuances but are computationally intensive, hindering real-time applications. In this paper, we present D2LLMs—Decomposed and Distilled LLMs for semantic search—that combines the best of both worlds. We decompose a cross-encoder into an efficient bi-encoder integrated with Pooling by Multihead Attention and an Interaction Emulation Module, achieving nuanced understanding and pre-computability. Knowledge from the LLM is distilled into this model using contrastive, rank, and feature imitation techniques. Our experiments show that D2LLM surpasses five leading baselines in terms of all metrics across three tasks, particularly improving NLI task performance by at least 6.45%

pdf bib
Can We Learn Question, Answer, and Distractors All from an Image? A New Task for Multiple-choice Visual Question Answering
Wenjian Ding | Yao Zhang | Jun Wang | Adam Jatowt | Zhenglu Yang
Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)

Multiple-choice visual question answering (MC VQA) requires an answer picked from a list of distractors, based on a question and an image. This research has attracted wide interest from the fields of visual question answering, visual question generation, and visual distractor generation. However, these fields still stay in their own territories, and how to jointly generate meaningful questions, correct answers, and challenging distractors remains unexplored. In this paper, we introduce a novel task, Visual Question-Answer-Distractors Generation (VQADG), which can bridge this research gap as well as take as a cornerstone to promote existing VQA models. Specific to the VQADG task, we present a novel framework consisting of a vision-and-language model to encode the given image and generate QADs jointly, and contrastive learning to ensure the consistency of the generated question, answer, and distractors. Empirical evaluations on the benchmark dataset validate the performance of our model in the VQADG task.

pdf bib
CHisIEC: An Information Extraction Corpus for Ancient Chinese History
Xuemei Tang | Qi Su | Jun Wang | Zekun Deng
Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)

Natural Language Processing (NLP) plays a pivotal role in the realm of Digital Humanities (DH) and serves as the cornerstone for advancing the structural analysis of historical and cultural heritage texts. This is particularly true for the domains of named entity recognition (NER) and relation extraction (RE). In our commitment to expediting ancient history and culture, we present the “Chinese Historical Information Extraction Corpus”(CHisIEC). CHisIEC is a meticulously curated dataset designed to develop and evaluate NER and RE tasks, offering a resource to facilitate research in the field. Spanning a remarkable historical timeline encompassing data from 13 dynasties spanning over 1830 years, CHisIEC epitomizes the extensive temporal range and text heterogeneity inherent in Chinese historical documents. The dataset encompasses four distinct entity types and twelve relation types, resulting in a meticulously labeled dataset comprising 14,194 entities and 8,609 relations. To establish the robustness and versatility of our dataset, we have undertaken comprehensive experimentation involving models of various sizes and paradigms. Additionally, we have evaluated the capabilities of Large Language Models (LLMs) in the context of tasks related to ancient Chinese history. The dataset and code are available at https://github.com/tangxuemei1995/CHisIEC.

pdf bib
Restoring Ancient Ideograph: A Multimodal Multitask Neural Network Approach
Siyu Duan | Jun Wang | Qi Su
Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)

Cultural heritage serves as the enduring record of human thought and history. Despite significant efforts dedicated to the preservation of cultural relics, many ancient artefacts have been ravaged irreversibly by natural deterioration and human actions. Deep learning technology has emerged as a valuable tool for restoring various kinds of cultural heritages, including ancient text restoration. Previous research has approached ancient text restoration from either visual or textual perspectives, often overlooking the potential of synergizing multimodal information. This paper proposes a novel Multimodal Multitask Restoring Model (MMRM) to restore ancient texts, particularly emphasising the ideograph. This model combines context understanding with residual visual information from damaged ancient artefacts, enabling it to predict damaged characters and generate restored images simultaneously. We tested the MMRM model through experiments conducted on both simulated datasets and authentic ancient inscriptions. The results show that the proposed method gives insightful restoration suggestions in both simulation experiments and real-world scenarios. To the best of our knowledge, this work represents the pioneering application of multimodal deep learning in ancient text restoration, which will contribute to the understanding of ancient society and culture in digital humanities fields.

2023

pdf bib
MGR: Multi-generator Based Rationalization
Wei Liu | Haozhao Wang | Jun Wang | Ruixuan Li | Xinyang Li | YuanKai Zhang | Yang Qiu
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Rationalization is to employ a generator and a predictor to construct a self-explaining NLP model in which the generator selects a subset of human-intelligible pieces of the input text to the following predictor. However, rationalization suffers from two key challenges, i.e., spurious correlation and degeneration, where the predictor overfits the spurious or meaningless pieces solely selected by the not-yet well-trained generator and in turn deteriorates the generator. Although many studies have been proposed to address the two challenges, they are usually designed separately and do not take both of them into account. In this paper, we propose a simple yet effective method named MGR to simultaneously solve the two problems. The key idea of MGR is to employ multiple generators such that the occurrence stability of real pieces is improved and more meaningful pieces are delivered to the predictor. Empirically, we show that MGR improves the F1 score by up to 20.9% as compared to state-of-the-art methods.

pdf bib
XSemPLR: Cross-Lingual Semantic Parsing in Multiple Natural Languages and Meaning Representations
Yusen Zhang | Jun Wang | Zhiguo Wang | Rui Zhang
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Cross-Lingual Semantic Parsing (CLSP) aims to translate queries in multiple natural languages (NLs) into meaning representations (MRs) such as SQL, lambda calculus, and logic forms. However, existing CLSP models are separately proposed and evaluated on datasets of limited tasks and applications, impeding a comprehensive and unified evaluation of CLSP on a diverse range of NLs and MRs. To this end, we present XSemPLR, a unified benchmark for cross-lingual semantic parsing featured with 22 natural languages and 8 meaning representations by examining and selecting 9 existing datasets to cover 5 tasks and 164 domains. We use XSemPLR to conduct a comprehensive benchmark study on a wide range of multilingual language models including encoder-based models (mBERT, XLM-R), encoder-decoder models (mBART, mT5), and decoder-based models (Codex, BLOOM). We design 6 experiment settings covering various lingual combinations (monolingual, multilingual, cross-lingual) and numbers of learning samples (full dataset, few-shot, and zero-shot). Our experiments show that encoder-decoder models (mT5) achieve the highest performance compared with other popular models, and multilingual training can further improve the average performance. Notably, multilingual large language models (e.g., BLOOM) are still inadequate to perform CLSP tasks. We also find that the performance gap between monolingual training and cross-lingual transfer learning is still significant for multilingual models, though it can be mitigated by cross-lingual few-shot training. Our dataset and code are available at https://github.com/psunlpgroup/XSemPLR.

pdf bib
Importance of Synthesizing High-quality Data for Text-to-SQL Parsing
Yiqun Hu | Yiyun Zhao | Jiarong Jiang | Wuwei Lan | Henghui Zhu | Anuj Chauhan | Alexander Hanbo Li | Lin Pan | Jun Wang | Chung-Wei Hang | Sheng Zhang | Jiang Guo | Mingwen Dong | Joseph Lilien | Patrick Ng | Zhiguo Wang | Vittorio Castelli | Bing Xiang
Findings of the Association for Computational Linguistics: ACL 2023

There has been increasing interest in synthesizing data to improve downstream text-to-SQL tasks. In this paper, we examined the existing synthesized datasets and discovered that state-of-the-art text-to-SQL algorithms did not further improve on popular benchmarks when trained with augmented synthetic data. We observed three shortcomings: illogical synthetic SQL queries from independent column sampling, arbitrary table joins, and language gaps between the synthesized SQL and natural language question (NLQ) pair. To address these issues, we propose a novel synthesis framework that imposes strong typing constraints, incorporates key relationships from schema, and conducts schema-distance-weighted column sampling. We also adopt an intermediate representation (IR) for the SQL-to-text task to further improve the quality of the generated NLQ. When existing powerful text-to-SQL parsers are pretrained on our high-quality synthesized data, these models have significant accuracy boosts and achieve new state-of-the-art performance on Spider. We also demonstrate the effectiveness of our techniques with ablation studies

pdf bib
ACROSS: An Alignment-based Framework for Low-Resource Many-to-One Cross-Lingual Summarization
Peiyao Li | Zhengkun Zhang | Jun Wang | Liang Li | Adam Jatowt | Zhenglu Yang
Findings of the Association for Computational Linguistics: ACL 2023

This research addresses the challenges of Cross-Lingual Summarization (CLS) in low-resource scenarios and over imbalanced multilingual data. Existing CLS studies mostly resort to pipeline frameworks or multi-task methods in bilingual settings. However, they ignore the data imbalance in multilingual scenarios and do not utilize the high-resource monolingual summarization data. In this paper, we propose the Aligned CROSs-lingual Summarization (ACROSS) model to tackle these issues. Our framework aligns low-resource cross-lingual data with high-resource monolingual data via contrastive and consistency loss, which help enrich low-resource information for high-quality summaries. In addition, we introduce a data augmentation method that can select informative monolingual sentences, which facilitates a deep exploration of high-resource information and introduce new information for low-resource languages. Experiments on the CrossSum dataset show that ACROSS outperforms baseline models and obtains consistently dominant performance on 45 language pairs.

pdf bib
Mitigating Backdoor Poisoning Attacks through the Lens of Spurious Correlation
Xuanli He | Qiongkai Xu | Jun Wang | Benjamin Rubinstein | Trevor Cohn
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

Modern NLP models are often trained over large untrusted datasets, raising the potential for a malicious adversary to compromise model behaviour. For instance, backdoors can be implanted through crafting training instances with a specific textual trigger and a target label. This paper posits that backdoor poisoning attacks exhibit a spurious correlation between simple text features and classification labels, and accordingly, proposes methods for mitigating spurious correlation as means of defence. Our empirical study reveals that the malicious triggers are highly correlated to their target labels; therefore such correlations are extremely distinguishable compared to those scores of benign features, and can be used to filter out potentially problematic instances. Compared with several existing defences, our defence method significantly reduces attack success rates across backdoor attacks, and in the case of insertion-based attacks, our method provides a near-perfect defence.

pdf bib
Well Begun is Half Done: Generator-agnostic Knowledge Pre-Selection for Knowledge-Grounded Dialogue
Lang Qin | Yao Zhang | Hongru Liang | Jun Wang | Zhenglu Yang
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

Accurate knowledge selection is critical in knowledge-grounded dialogue systems. Towards a closer look at it, we offer a novel perspective to organize existing literature, i.e., knowledge selection coupled with, after, and before generation. We focus on the third under-explored category of study, which can not only select knowledge accurately in advance, but has the advantage to reduce the learning, adjustment, and interpretation burden of subsequent response generation models, especially LLMs. We propose \tt{GATE}, a generator-agnostic knowledge selection method, to prepare knowledge for subsequent response generation models by selecting context-related knowledge among different knowledge structures and variable knowledge requirements. Experimental results demonstrate the superiority of \tt{GATE}, and indicate that knowledge selection before generation is a lightweight yet effective way to facilitate LLMs (e.g., ChatGPT) to generate more informative responses.

pdf bib
How Do Large Language Models Capture the Ever-changing World Knowledge? A Review of Recent Advances
Zihan Zhang | Meng Fang | Ling Chen | Mohammad-Reza Namazi-Rad | Jun Wang
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

Although large language models (LLMs) are impressive in solving various tasks, they can quickly be outdated after deployment. Maintaining their up-to-date status is a pressing concern in the current era. This paper provides a comprehensive review of recent advances in aligning deployed LLMs with the ever-changing world knowledge. We categorize research works systemically and provide in-depth comparisons and discussions. We also discuss existing challenges and highlight future directions to facilitate research in this field.

pdf bib
CCL23-Eval任务1总结报告:古籍命名实体识别(GuNER2023)(Overview of CCL23-Eval Task 1: Named Entity Recognition in Ancient Chinese Books)
Qi Su (祺苏,) | Yingying Wang (王莹莹) | Zekun Deng (邓泽琨) | Hao Yang (杨浩) | Jun Wang (王军)
Proceedings of the 22nd Chinese National Conference on Computational Linguistics (Volume 3: Evaluations)

“第23届中国计算语言学大会(CCL)提出了中文信息处理方面的10个评测任务。其中,任务1为古籍命名实体识别评测,由北京大学数字人文研究中心、北京大学人工智能研究院组织。该任务的主要目标是自动识别古籍文本中事件基本构成要素的重要实体,以提供对古汉语文本进行分析处理的基础。评测发布了覆盖多个朝代和领域的”二十四史”评测数据集,共15万余字,包含人名、书名、官职名三种实体超万数。同时设置了封闭和开放两个赛道,聚焦于不同规格的预训练模型的应用能力。共有127支队伍报名参加了该评测任务。在封闭赛道上,参赛系统在测试集上的最佳性能达到了96.15%的F1值;在开放赛道上,最佳性能达到了95.48%的F1值。”

pdf bib
IMBERT: Making BERT Immune to Insertion-based Backdoor Attacks
Xuanli He | Jun Wang | Benjamin Rubinstein | Trevor Cohn
Proceedings of the 3rd Workshop on Trustworthy Natural Language Processing (TrustNLP 2023)

Backdoor attacks are an insidious security threat against machine learning models. Adversaries can manipulate the predictions of compromised models by inserting triggers into the training phase. Various backdoor attacks have been devised which can achieve nearly perfect attack success without affecting model predictions for clean inputs. Means of mitigating such vulnerabilities are underdeveloped, especially in natural language processing. To fill this gap, we introduce IMBERT, which uses either gradients or self-attention scores derived from victim models to self-defend against backdoor attacks at inference time. Our empirical studies demonstrate that IMBERT can effectively identify up to 98.5% of inserted triggers. Thus, it significantly reduces the attack success rate while attaining competitive accuracy on the clean dataset across widespread insertion-based attacks compared to two baselines. Finally, we show that our approach is model-agnostic, and can be easily ported to several pre-trained transformer models.

2022

pdf bib
Multi-Party Empathetic Dialogue Generation: A New Task for Dialog Systems
Ling.Yu Zhu | Zhengkun Zhang | Jun Wang | Hongbin Wang | Haiying Wu | Zhenglu Yang
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Empathetic dialogue assembles emotion understanding, feeling projection, and appropriate response generation. Existing work for empathetic dialogue generation concentrates on the two-party conversation scenario. Multi-party dialogues, however, are pervasive in reality. Furthermore, emotion and sensibility are typically confused; a refined empathy analysis is needed for comprehending fragile and nuanced human feelings. We address these issues by proposing a novel task called Multi-Party Empathetic Dialogue Generation in this study. Additionally, a Static-Dynamic model for Multi-Party Empathetic Dialogue Generation, SDMPED, is introduced as a baseline by exploring the static sensibility and dynamic emotion for the multi-party empathetic dialogue learning, the aspects that help SDMPED achieve the state-of-the-art performance.

pdf bib
Cross-Utterance Conditioned VAE for Non-Autoregressive Text-to-Speech
Yang Li | Cheng Yu | Guangzhi Sun | Hua Jiang | Fanglei Sun | Weiqin Zu | Ying Wen | Yang Yang | Jun Wang
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Modelling prosody variation is critical for synthesizing natural and expressive speech in end-to-end text-to-speech (TTS) systems. In this paper, a cross-utterance conditional VAE (CUC-VAE) is proposed to estimate a posterior probability distribution of the latent prosody features for each phoneme by conditioning on acoustic features, speaker information, and text features obtained from both past and future sentences. At inference time, instead of the standard Gaussian distribution used by VAE, CUC-VAE allows sampling from an utterance-specific prior distribution conditioned on cross-utterance information, which allows the prosody features generated by the TTS system to be related to the context and is more similar to how humans naturally produce prosody. The performance of CUC-VAE is evaluated via a qualitative listening test for naturalness, intelligibility and quantitative measurements, including word error rates and the standard deviation of prosody attributes. Experimental results on LJ-Speech and LibriTTS data show that the proposed CUC-VAE TTS system improves naturalness and prosody diversity with clear margins.

pdf bib
Modeling Temporal-Modal Entity Graph for Procedural Multimodal Machine Comprehension
Huibin Zhang | Zhengkun Zhang | Yao Zhang | Jun Wang | Yufan Li | Ning Jiang | Xin Wei | Zhenglu Yang
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Procedural Multimodal Documents (PMDs) organize textual instructions and corresponding images step by step. Comprehending PMDs and inducing their representations for the downstream reasoning tasks is designated as Procedural MultiModal Machine Comprehension (M3C). In this study, we approach Procedural M3C at a fine-grained level (compared with existing explorations at a document or sentence level), that is, entity. With delicate consideration, we model entity both in its temporal and cross-modal relation and propose a novel Temporal-Modal Entity Graph (TMEG). Specifically, graph structure is formulated to capture textual and visual entities and trace their temporal-modal evolution. In addition, a graph aggregation module is introduced to conduct graph encoding and reasoning. Comprehensive experiments across three Procedural M3C tasks are conducted on a traditional dataset RecipeQA and our new dataset CraftQA, which can better evaluate the generalization of TMEG.

pdf bib
Measuring and Mitigating Name Biases in Neural Machine Translation
Jun Wang | Benjamin Rubinstein | Trevor Cohn
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Neural Machine Translation (NMT) systems exhibit problematic biases, such as stereotypical gender bias in the translation of occupation terms into languages with grammatical gender. In this paper we describe a new source of bias prevalent in NMT systems, relating to translations of sentences containing person names. To correctly translate such sentences, a NMT system needs to determine the gender of the name. We show that leading systems are particularly poor at this task, especially for female given names. This bias is deeper than given name gender: we show that the translation of terms with ambiguous sentiment can also be affected by person names, and the same holds true for proper nouns denoting race. To mitigate these biases we propose a simple but effective data augmentation method based on randomly switching entities during translation, which effectively eliminates the problem without any effect on translation quality.

pdf bib
Improving Text-to-SQL Semantic Parsing with Fine-grained Query Understanding
Jun Wang | Patrick Ng | Alexander Hanbo Li | Jiarong Jiang | Zhiguo Wang | Bing Xiang | Ramesh Nallapati | Sudipta Sengupta
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: Industry Track

Most recent research on Text-to-SQL semantic parsing relies on either parser itself or simple heuristic based approach to understand natural language query (NLQ). When synthesizing a SQL query, there is no explicit semantic information of NLQ available to the parser which leads to undesirable generalization performance. In addition, without lexical-level fine-grained query understanding, linking between query and database can only rely on fuzzy string match which leads to suboptimal performance in real applications. In view of this, in this paper we present a general-purpose, modular neural semantic parsing framework that is based on token-level fine-grained query understanding. Our framework consists of three modules: named entity recognizer (NER), neural entity linker (NEL) and neural semantic parser (NSP). By jointly modeling query and database, NER model analyzes user intents and identifies entities in the query. NEL model links typed entities to schema and cell values in database. Parser model leverages available semantic information and linking results and synthesizes tree-structured SQL queries based on dynamically generated grammar. Experiments on SQUALL, a newly released semantic parsing dataset, show that we can achieve 56.8% execution accuracy on WikiTableQuestions (WTQ) test set, which outperforms the state-of-the-art model by 2.7%.

pdf bib
Foiling Training-Time Attacks on Neural Machine Translation Systems
Jun Wang | Xuanli He | Benjamin Rubinstein | Trevor Cohn
Findings of the Association for Computational Linguistics: EMNLP 2022

Neural machine translation (NMT) systems are vulnerable to backdoor attacks, whereby an attacker injects poisoned samples into training such that a trained model produces malicious translations. Nevertheless, there is little research on defending against such backdoor attacks in NMT. In this paper, we first show that backdoor attacks that have been successful in text classification are also effective against machine translation tasks. We then present a novel defence method that exploits a key property of most backdoor attacks: namely the asymmetry between the source and target language sentences, which is used to facilitate malicious text insertions, substitutions and suchlike. Our technique uses word alignment coupled with language model scoring to detect outlier tokens, and thus can find and filter out training instances which may contain backdoors. Experimental results demonstrate that our technique can significantly reduce the success of various attacks by up to 89.0%, while not affecting predictive accuracy.

pdf bib
数字人文视角下的《史记》《汉书》比较研究(A Comparative Study of Shiji and Hanshu from the Perspective of Digital Humanities)
Zekun Deng (邓泽琨) | Hao Yang (杨浩) | Jun Wang (王军)
Proceedings of the 21st Chinese National Conference on Computational Linguistics

“《史记》和《汉书》具有经久不衰的研究价值。尽管两书异同的研究已经较为丰富,但研究的全面性、完备性、科学性、客观性均仍显不足。在数字人文的视角下,本文利用计算语言学方法,通过对字、词、命名实体、段落等的多粒度、多角度分析,开展对于《史》《汉》的比较研究。首先,本文对于《史》《汉》中的字、词、命名实体的分布和特点进行对比,以遍历穷举的考察方式提炼出两书在主要内容上的相同点与不同点,揭示了汉武帝之前和汉武帝到西汉灭亡两段历史时期在政治、文化、思想上的重要变革与承袭。其次,本文使用一种融入命名实体作为外部特征的文本相似度算法对于《史记》《汉书》的异文进行自动发现,成功识别出过去研究者通过人工手段没有发现的袭用段落,使得我们对于《史》《汉》的承袭关系形成更加完整和立体的认识。再次,本文通过计算异文段落之间的最长公共子序列来自动得出两段异文之间存在的差异,从宏观统计上证明了《汉书》文字风格《史记》的差别,并从微观上进一步对二者语言特点进行了阐释,为理解《史》《汉》异文特点提供了新的角度和启发。本研究站在数字人文的视域下,利用先进的计算方法对于传世千年的中国古代经典进行了再审视、再发现,其方法对于今人研究古籍有一定的借鉴价值。”

pdf bib
PA Ph&Tech at SemEval-2022 Task 11: NER Task with Ensemble Embedding from Reinforcement Learning
Qizhi Lin | Changyu Hou | Xiaopeng Wang | Jun Wang | Yixuan Qiao | Peng Jiang | Xiandi Jiang | Benqi Wang | Qifeng Xiao
Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)

From pretrained contextual embedding to document-level embedding, the selection and construction of embedding have drawn more and more attention in the NER domain in recent research. This paper aims to discuss the performance of ensemble embeddings on complex NER tasks. Enlightened by Wang’s methodology, we try to replicate the dominating power of ensemble models with reinforcement learning optimizor on plain NER tasks to complex ones. Based on the composition of semeval dataset, the performance of the applied model is tested on lower-context, QA, and search query scenarios together with its zero-shot learning ability. Results show that with abundant training data, the model can achieve similar performance on lower-context cases compared to plain NER cases, but can barely transfer the performance to other scenarios in the test phase.

pdf bib
SFE-AI at SemEval-2022 Task 11: Low-Resource Named Entity Recognition using Large Pre-trained Language Models
Changyu Hou | Jun Wang | Yixuan Qiao | Peng Jiang | Peng Gao | Guotong Xie | Qizhi Lin | Xiaopeng Wang | Xiandi Jiang | Benqi Wang | Qifeng Xiao
Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)

Large scale pre-training models have been widely used in named entity recognition (NER) tasks. However, model ensemble through parameter averaging or voting can not give full play to the differentiation advantages of different models, especially in the open domain. This paper describes our NER system in the SemEval 2022 task11: MultiCoNER. We proposed an effective system to adaptively ensemble pre-trained language models by a Transformer layer. By assigning different weights to each model for different inputs, we adopted the Transformer layer to integrate the advantages of diverse models effectively. Experimental results show that our method achieves superior performances in Farsi and Dutch.

pdf bib
Eureka: Neural Insight Learning for Knowledge Graph Reasoning
Alex X. Zhang | Xun Liang | Bo Wu | Xiangping Zheng | Sensen Zhang | Yuhui Guo | Jun Wang | Xinyao Liu
Proceedings of the 29th International Conference on Computational Linguistics

The human recognition system has presented the remarkable ability to effortlessly learn novel knowledge from only a few trigger events based on prior knowledge, which is called insight learning. Mimicking such behavior on Knowledge Graph Reasoning (KGR) is an interesting and challenging research problem with many practical applications. Simultaneously, existing works, such as knowledge embedding and few-shot learning models, have been limited to conducting KGR in either “seen-to-seen” or “unseen-to-unseen” scenarios. To this end, we propose a neural insight learning framework named Eureka to bridge the “seen” to “unseen” gap. Eureka is empowered to learn the seen relations with sufficient training triples while providing the flexibility of learning unseen relations given only one trigger without sacrificing its performance on seen relations. Eureka meets our expectation of the model to acquire seen and unseen relations at no extra cost, and eliminate the need to retrain when encountering emerging unseen relations. Experimental results on two real-world datasets demonstrate that the proposed framework also outperforms various state-of-the-art baselines on datasets of both seen and unseen relations.

2021

pdf bib
Self Promotion in US Congressional Tweets
Jun Wang | Kelly Cui | Bei Yu
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies

Prior studies have found that women self-promote less than men due to gender stereotypes. In this study we built a BERT-based NLP model to predict whether a Congressional tweet shows self-promotion or not and then used this model to examine whether a gender gap in self-promotion exists among Congressional tweets. After analyzing 2 million Congressional tweets from July 2017 to March 2021, controlling for a number of factors that include political party, chamber, age, number of terms in Congress, number of daily tweets, and number of followers, we found that women in Congress actually perform more self-promotion on Twitter, indicating a reversal of traditional gender norms where women self-promote less than men.

pdf bib
MM-AVS: A Full-Scale Dataset for Multi-modal Summarization
Xiyan Fu | Jun Wang | Zhenglu Yang
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies

Multimodal summarization becomes increasingly significant as it is the basis for question answering, Web search, and many other downstream tasks. However, its learning materials have been lacking a holistic organization by integrating resources from various modalities, thereby lagging behind the research progress of this field. In this study, we release a full-scale multimodal dataset comprehensively gathering documents, summaries, images, captions, videos, audios, transcripts, and titles in English from CNN and Daily Mail. To our best knowledge, this is the first collection that spans all modalities and nearly comprises all types of materials available in this community. In addition, we devise a baseline model based on the novel dataset, which employs a newly proposed Jump-Attention mechanism based on transcripts. The experimental results validate the important assistance role of the external information for multimodal summarization.

pdf bib
基于预训练语言模型的繁体古文自动句读研究(Automatic Traditional Ancient Chinese Texts Segmentation and Punctuation Based on Pre-training Language Model)
Xuemei Tang (唐雪梅) | Qi Su (苏祺) | Jun Wang (王军) | Yuhang Chen (陈雨航) | Hao Yang (杨浩)
Proceedings of the 20th Chinese National Conference on Computational Linguistics

未经整理的古代典籍不含任何标点,不符合当代人的阅读习惯,古籍断句标点之后有助于阅读、研究和出版。本文提出了一种基于预训练语言模型的繁体古文自动句读框架。本文整理了约10亿字的繁体古文语料,对于训练语言模型进行增量训练,在此基础上上实现古文自动句读和标点。实验表明经过大规模繁体古文语料增量训练后的语言模型具备更好的古文语义表示能力,能够有助提升繁体古文自动句读和自动标点的效果。融合了增量训练模型之后,古文断句F1值达到95.03%,古文标点F1值达到了80.18%,分别比使用未增量训练的语言模型提升1.83%和2.21%。为解决现有篇章级句读方案效率低的问题,本文改进了前人的串行滑动窗口方式,在一定程度上提高了句读效率,并提出一种新的并行滑动窗口方式,能够高效准确地进行长文本自动句读。

pdf bib
Putting words into the system’s mouth: A targeted attack on neural machine translation using monolingual data poisoning
Jun Wang | Chang Xu | Francisco Guzmán | Ahmed El-Kishky | Yuqing Tang | Benjamin Rubinstein | Trevor Cohn
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021

pdf bib
As Easy as 1, 2, 3: Behavioural Testing of NMT Systems for Numerical Translation
Jun Wang | Chang Xu | Francisco Guzmán | Ahmed El-Kishky | Benjamin Rubinstein | Trevor Cohn
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021

pdf bib
Mitigating Data Poisoning in Text Classification with Differential Privacy
Chang Xu | Jun Wang | Francisco Guzmán | Benjamin Rubinstein | Trevor Cohn
Findings of the Association for Computational Linguistics: EMNLP 2021

NLP models are vulnerable to data poisoning attacks. One type of attack can plant a backdoor in a model by injecting poisoned examples in training, causing the victim model to misclassify test instances which include a specific pattern. Although defences exist to counter these attacks, they are specific to an attack type or pattern. In this paper, we propose a generic defence mechanism by making the training process robust to poisoning attacks through gradient shaping methods, based on differentially private training. We show that our method is highly effective in mitigating, or even eliminating, poisoning attacks on text classification, with only a small cost in predictive accuracy.

pdf bib
Detecting Health Advice in Medical Research Literature
Yingya Li | Jun Wang | Bei Yu
Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

Health and medical researchers often give clinical and policy recommendations to inform health practice and public health policy. However, no current health information system supports the direct retrieval of health advice. This study fills the gap by developing and validating an NLP-based prediction model for identifying health advice in research publications. We annotated a corpus of 6,000 sentences extracted from structured abstracts in PubMed publications as ‘“strong advice”, “weak advice”, or “no advice”, and developed a BERT-based model that can predict, with a macro-averaged F1-score of 0.93, whether a sentence gives strong advice, weak advice, or not. The prediction model generalized well to sentences in both unstructured abstracts and discussion sections, where health advice normally appears. We also conducted a case study that applied this prediction model to retrieve specific health advice on COVID-19 treatments from LitCovid, a large COVID research literature portal, demonstrating the usefulness of retrieving health advice sentences as an advanced research literature navigation function for health researchers and the general public.

2020

pdf bib
Measuring Correlation-to-Causation Exaggeration in Press Releases
Bei Yu | Jun Wang | Lu Guo | Yingya Li
Proceedings of the 28th International Conference on Computational Linguistics

Press releases have an increasingly strong influence on media coverage of health research; however, they have been found to contain seriously exaggerated claims that can misinform the public and undermine public trust in science. In this study we propose an NLP approach to identify exaggerated causal claims made in health press releases that report on observational studies, which are designed to establish correlational findings, but are often exaggerated as causal. We developed a new corpus and trained models that can identify causal claims in the main statements in a press release. By comparing the claims made in a press release with the corresponding claims in the original research paper, we found that 22% of press releases made exaggerated causal claims from correlational findings in observational studies. Furthermore, universities exaggerated more often than journal publishers by a ratio of 1.5 to 1. Encouragingly, the exaggeration rate has slightly decreased over the past 10 years, despite the increase of the total number of press releases. More research is needed to understand the cause of the decreasing pattern.

pdf bib
Diversify Question Generation with Continuous Content Selectors and Question Type Modeling
Zhen Wang | Siwei Rao | Jie Zhang | Zhen Qin | Guangjian Tian | Jun Wang
Findings of the Association for Computational Linguistics: EMNLP 2020

Generating questions based on answers and relevant contexts is a challenging task. Recent work mainly pays attention to the quality of a single generated question. However, question generation is actually a one-to-many problem, as it is possible to raise questions with different focuses on contexts and various means of expression. In this paper, we explore the diversity of question generation and come up with methods from these two aspects. Specifically, we relate contextual focuses with content selectors, which are modeled by a continuous latent variable with the technique of conditional variational auto-encoder (CVAE). In the realization of CVAE, a multimodal prior distribution is adopted to allow for more diverse content selectors. To take into account various means of expression, question types are explicitly modeled and a diversity-promoting algorithm is proposed further. Experimental results on public datasets show that our proposed method can significantly improve the diversity of generated questions, especially from the perspective of using different question types. Overall, our proposed method achieves a better trade-off between generation quality and diversity compared with existing approaches.

2019

pdf bib
Detecting Causal Language Use in Science Findings
Bei Yu | Yingya Li | Jun Wang
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)

Causal interpretation of correlational findings from observational studies has been a major type of misinformation in science communication. Prior studies on identifying inappropriate use of causal language relied on manual content analysis, which is not scalable for examining a large volume of science publications. In this study, we first annotated a corpus of over 3,000 PubMed research conclusion sentences, then developed a BERT-based prediction model that classifies conclusion sentences into “no relationship”, “correlational”, “conditional causal”, and “direct causal” categories, achieving an accuracy of 0.90 and a macro-F1 of 0.88. We then applied the prediction model to measure the causal language use in the research conclusions of about 38,000 observational studies in PubMed. The prediction result shows that 21.7% studies used direct causal language exclusively in their conclusions, and 32.4% used some direct causal language. We also found that the ratio of causal language use differs among authors from different countries, challenging the notion of a shared consensus on causal language use in the global science community. Our prediction model could also be used to help identify the inappropriate use of causal language in science publications.

pdf bib
Permanent Magnetic Articulograph (PMA) vs Electromagnetic Articulograph (EMA) in Articulation-to-Speech Synthesis for Silent Speech Interface
Beiming Cao | Nordine Sebkhi | Ted Mau | Omer T. Inan | Jun Wang
Proceedings of the Eighth Workshop on Speech and Language Processing for Assistive Technologies

Silent speech interfaces (SSIs) are devices that enable speech communication when audible speech is unavailable. Articulation-to-speech (ATS) synthesis is a software design in SSI that directly converts articulatory movement information into audible speech signals. Permanent magnetic articulograph (PMA) is a wireless articulator motion tracking technology that is similar to commercial, wired Electromagnetic Articulograph (EMA). PMA has shown great potential for practical SSI applications, because it is wireless. The ATS performance of PMA, however, is unknown when compared with current EMA. In this study, we compared the performance of ATS using a PMA we recently developed and a commercially available EMA (NDI Wave system). Datasets with same stimuli and size that were collected from tongue tip were used in the comparison. The experimental results indicated the performance of PMA was close to, although not as equally good as that of EMA. Furthermore, in PMA, converting the raw magnetic signals to positional signals did not significantly affect the performance of ATS, which support the future direction in PMA-based ATS can be focused on the use of positional signals to maximize the benefit of spatial analysis.

pdf bib
Speech-based Estimation of Bulbar Regression in Amyotrophic Lateral Sclerosis
Alan Wisler | Kristin Teplansky | Jordan Green | Yana Yunusova | Thomas Campbell | Daragh Heitzman | Jun Wang
Proceedings of the Eighth Workshop on Speech and Language Processing for Assistive Technologies

Amyotrophic Lateral Sclerosis (ALS) is a progressive neurological disease that leads to degeneration of motor neurons and, as a result, inhibits the ability of the brain to control muscle movements. Monitoring the progression of ALS is of fundamental importance due to the wide variability in disease outlook that exists across patients. This progression is typically tracked using the ALS functional rating scale - revised (ALSFRS-R), which is the current clinical assessment of a patient’s level of functional impairment including speech and other motor tasks. In this paper, we investigated automatic estimation of the ALSFRS-R bulbar subscore from acoustic and articulatory movement samples. Experimental results demonstrated the AFSFRS-R bulbar subscore can be predicted from speech samples, which has clinical implication for automatic monitoring of the disease progression of ALS using speech information.

2018

pdf bib
JTAV: Jointly Learning Social Media Content Representation by Fusing Textual, Acoustic, and Visual Features
Hongru Liang | Haozheng Wang | Jun Wang | Shaodi You | Zhe Sun | Jin-Mao Wei | Zhenglu Yang
Proceedings of the 27th International Conference on Computational Linguistics

Learning social media content is the basis of many real-world applications, including information retrieval and recommendation systems, among others. In contrast with previous works that focus mainly on single modal or bi-modal learning, we propose to learn social media content by fusing jointly textual, acoustic, and visual information (JTAV). Effective strategies are proposed to extract fine-grained features of each modality, that is, attBiGRU and DCRNN. We also introduce cross-modal fusion and attentive pooling techniques to integrate multi-modal information comprehensively. Extensive experimental evaluation conducted on real-world datasets demonstrate our proposed model outperforms the state-of-the-art approaches by a large margin.

2015

pdf bib
Recognizing Dysarthric Speech due to Amyotrophic Lateral Sclerosis with Across-Speaker Articulatory Normalization
Seongjun Hahm | Daragh Heitzman | Jun Wang
Proceedings of SLPAT 2015: 6th Workshop on Speech and Language Processing for Assistive Technologies

pdf bib
Determining an Optimal Set of Flesh Points on Tongue, Lips, and Jaw for Continuous Silent Speech Recognition
Jun Wang | Seongjun Hahm | Ted Mau
Proceedings of SLPAT 2015: 6th Workshop on Speech and Language Processing for Assistive Technologies

2014

pdf bib
Preliminary Test of a Real-Time, Interactive Silent Speech Interface Based on Electromagnetic Articulograph
Jun Wang | Ashok Samal | Jordan Green
Proceedings of the 5th Workshop on Speech and Language Processing for Assistive Technologies

2013

pdf bib
Word Recognition from Continuous Articulatory Movement Time-series Data using Symbolic Representations
Jun Wang | Arvind Balasubramanian | Luis Mojica de la Vega | Jordan R. Green | Ashok Samal | Balakrishnan Prabhakaran
Proceedings of the Fourth Workshop on Speech and Language Processing for Assistive Technologies

2008

pdf bib
A Data Driven Approach to Query Expansion in Question Answering
Leon Derczynski | Jun Wang | Robert Gaizauskas | Mark A. Greenwood
Coling 2008: Proceedings of the 2nd workshop on Information Retrieval for Question Answering

2003

pdf bib
On Intra-page and Inter-page Semantic Analysis of Web Pages
Jun Wang | Jicheng Wang | Gangshan Wu | Hiroshi Tsuda
Proceedings of the 17th Pacific Asia Conference on Language, Information and Computation

Search
Co-authors