Pengyuan Liu

Also published as: Peng-Yuan Liu, PengYuan Liu


2024

pdf bib
Do PLMs and Annotators Share the Same Gender Bias? Definition, Dataset, and Framework of Contextualized Gender Bias
Shucheng Zhu | Bingjie Du | Jishun Zhao | Ying Liu | Pengyuan Liu
Proceedings of the 5th Workshop on Gender Bias in Natural Language Processing (GeBNLP)

Pre-trained language models (PLMs) have achieved success in various of natural language processing (NLP) tasks. However, PLMs also introduce some disquieting safety problems, such as gender bias. Gender bias is an extremely complex issue, because different individuals may hold disparate opinions on whether the same sentence expresses harmful bias, especially those seemingly neutral or positive. This paper first defines the concept of contextualized gender bias (CGB), which makes it easy to measure implicit gender bias in both PLMs and annotators. We then construct CGBDataset, which contains 20k natural sentences with gendered words, from Chinese news. Similar to the task of masked language models, gendered words are masked for PLMs and annotators to judge whether a male word or a female word is more suitable. Then, we introduce CGBFrame to measure the gender bias of annotators. By comparing the results measured by PLMs and annotators, we find that though there are differences on the choices made by PLMs and annotators, they show significant consistency in general.

pdf
MatPlotAgent: Method and Evaluation for LLM-Based Agentic Scientific Data Visualization
Zhiyu Yang | Zihan Zhou | Shuo Wang | Xin Cong | Xu Han | Yukun Yan | Zhenghao Liu | Zhixing Tan | Pengyuan Liu | Dong Yu | Zhiyuan Liu | Xiaodong Shi | Maosong Sun
Findings of the Association for Computational Linguistics ACL 2024

Scientific data visualization plays a crucial role in research by enabling the direct display of complex information and assisting researchers in identifying implicit patterns. Despite its importance, the use of Large Language Models (LLMs) for scientific data visualization remains rather unexplored. In this study, we introduce MatPlotAgent, an efficient model-agnostic LLM agent framework designed to automate scientific data visualization tasks. Leveraging the capabilities of both code LLMs and multi-modal LLMs, MatPlotAgent consists of three core modules: query understanding, code generation with iterative debugging, and a visual feedback mechanism for error correction. To address the lack of benchmarks in this field, we present MatPlotBench, a high-quality benchmark consisting of 100 human-verified test cases. Additionally, we introduce a scoring approach that utilizes GPT-4V for automatic evaluation. Experimental results demonstrate that MatPlotAgent can improve the performance of various LLMs, including both commercial and open-source models. Furthermore, the proposed evaluation method shows a strong correlation with human-annotated scores.

2023

pdf
中国社会道德变化模型与发展动因探究——基于70年《人民日报》的计量与分析 (The Model of Moral Change and Motivation in Chinese Society ——The Vocabulary Analysis of the 70-year ”People’s Daily”)
Hongrui Wang (王弘睿) | Dong Yu (于东) | Pengyuan Liu (刘鹏远) | Liying Ceng (曾立英)
Proceedings of the 22nd Chinese National Conference on Computational Linguistics

“社会道德的历时变迁研究具有重要意义。通过观察语言使用与道德变迁的历时联系,能够帮助描绘社会道德的变化趋势和发展规律、把握社会道德动态、推进道德建设。目前缺少从词汇角度、利用计算手段对大规模历时语料进行系统、全面的社会道德变迁研究。基于此,该文提出道德主题词历时计量模型,通过计量指标对1946-2015共70年的《人民日报》语料进行了历时计算与分析,观察了70年社会道德主题词的使用选择与变化。研究结果发现,道德词汇的历时使用与社会道德之间存在互动关系,反映出70年中国社会道德的历时变革与发展情况。”

pdf
动词视角下的汉语性别表征研究——基于多语体语料库与依存分析(Gendered Representation in Chinese via Verbal Analysis —Based on a Multi-register Corpus and Dependency Parsing)
Yingshi Chen (陈颖诗) | Dong Yu (于东) | Pengyuan Liu (刘鹏远)
Proceedings of the 22nd Chinese National Conference on Computational Linguistics

“动作是反映性别社会化的重要形式,研究汉语中动词的性别表征,可以找到语言构建不同性别身份的路径,即所采用的方式、形式。本文以依存句法关系为抓手,在四种语体的语料中抽取出和不同性别词构成依存结构的动词,统计出有显著性别差异的动词,并根据性别词充当的句子成分,结合语义进行了定量和定性分析。总体来看,大部分汉语动词表征是中性的,能体现性别的动词是少数,汉语作为一种承载着中华智慧且具有深厚文化底蕴的语言,对性别的表征是中立且平等的,这也体现出了我国的性别平等观念。而在表征性别的动词中,能看到构建男性和女性身份的两种不同路径。显著表征女性的动词在不同语体的语料中均多于显著表征男性的,但是表征男性的动词的语义分布则更为均衡,体现了“男性默认-女性专门”。在司法动词上,女性常常作为暴力行为的受害者,同时施害者男性却隐身了,体现了筜男性主宰笭女性顺从笢。不同语体的动词在构建性别时体现了不同的功能,新闻塑造了较为传统的性别规范,传统和网络文学以不同的形式打破了固有的性别规范。”

pdf
大规模语言模型增强的中文篇章多维度阅读体验量化研究(Quantitative Research on Multi-dimensional Reading Experience of Chinese Texts Enhanced by Large Language Model)
Jiadai Sun (孙嘉黛) | Siyi Tang (汤思怡) | Shike Wang (王诗可) | Dong Yu (于东) | Pengyuan Liu (刘鹏远)
Proceedings of the 22nd Chinese National Conference on Computational Linguistics

“现有的文本分级阅读研究往往从文本可读性的角度出发,以离散的文本难度等级的形式为读者推荐阅读书目。目前,仍缺少一种研究读者在阅读过程中产生的多方面、深层次阅读体验的体系结构。对此,我们调研了读者在阅读中文篇章过程中产生的不同阅读体验,提出了中文篇章多维度阅读体验的量化体系。我们将阅读过程中呈现的连续性的阅读体验归纳为多种类别,并在此基础上构建了中文篇章多维度阅读体验数据集。同时,我们探究了以大规模语言模型为基础的ChatGPT对阅读体验的量化能力,发现其虽具备强大的信息抽取和语义理解能力,在阅读体验的量化上却表现不佳。但我们发现大规模语言模型所蕴含的能力能够以知识蒸馏的方式协助深层属性的量化,基于此,我们实现了大规模语言模型增强的中文篇章多维阅读体验量化模型。模型在各维度阅读体验上的平均F1值达到0.72,高于ChatGPT的Fewshot结果0.48。”

2022

pdf
CoreValue:面向价值观计算的中文核心价值-行为体系及知识库(CoreValue: Chinese Core Value-Behavior Frame and Knowledge Base for Value Computing)
Pengyuan Liu (刘鹏远) | Sanle Zhang (张三乐) | Dong Yu (于东) | Lin Bo (薄琳)
Proceedings of the 21st Chinese National Conference on Computational Linguistics

“由主体行为推断其价值观是人工智能理解并具有人类价值观的前提之一。在NLP相关领域,研究主要集中在对文本价值观或道德的是非判断上,鲜见由主体行为推断其价值观的工作,也缺乏相应的数据资源。该文首先构建了中文核心价值-行为体系。该体系以社会主义核心价值观为基础,分为两部分:1)类别体系。共包含8大类核心价值,进一步细分为19小类双方向价值并对应38类行为;2)要素体系。划分为核心与非核心要素共7种。随后,抽取语料中含有主体行为的文本句,依据该体系进行人工标注,构建了一个包含6994个行为句及其对应的细粒度价值与方向,34965个要素的细粒度中文价值-行为知识库。最后,该文提出了价值观类别判别、方向判别及联合判别任务并进行了实验。结果表明,基于预训练语言模型的方法在价值观方向判别上表现优异,在细粒度价值类别判别以及价值类别多标签判别上,有较大提升空间。”

pdf
中文自然语言处理多任务中的职业性别偏见测量(Measurement of Occupational Gender Bias in Chinese Natural Language Processing Tasks)
Mengqing Guo (郭梦清) | Jiali Li (李加厉) | Jishun Zhao (赵继舜) | Shucheng Zhu (朱述承) | Ying Liu (刘颖) | Pengyuan Liu (刘鹏远)
Proceedings of the 21st Chinese National Conference on Computational Linguistics

“尽管悲观者认为,职场中永远不可能存在性别平等。但随着人们观念的转变,愈来愈多的人们相信,职业的选择应只与个人能力相匹配,而不应由个体的性别决定。目前已经发现自然语言处理的各个任务中都存在着职业性别偏见。但这些研究往往只针对特定的英文任务,缺乏针对中文的、综合多任务的职业性别偏见测量研究。本文基于霍兰德职业模型,从中文自然语言处理中常见的三个任务出发,测量了词向量、共指消解和文本生成中的职业性别偏见,发现不同任务中的职业性别偏见既有一定的共性,又存在着独特的差异性。总体来看,不同任务中的职业性别偏见反映了现实生活中人们对于不同性别所选择职业的刻板印象。此外,在设计不同任务的偏见测量指标时,还需要考虑如语体、词序等语言学要素的影响。”

pdf
From Polarity to Intensity: Mining Morality from Semantic Space
Chunxu Zhao | Pengyuan Liu | Dong Yu
Proceedings of the 29th International Conference on Computational Linguistics

Most works on computational morality focus on moral polarity recognition, i.e., distinguishing right from wrong. However, a discrete polarity label is not informative enough to reflect morality as it does not contain any degree or intensity information. Existing approaches to compute moral intensity are limited to word-level measurement and heavily rely on human labelling. In this paper, we propose MoralScore, a weakly-supervised framework that can automatically measure moral intensity from text. It only needs moral polarity labels, which are more robust and easier to acquire. Besides, the framework can capture latent moral information not only from words but also from sentence-level semantics which can provide a more comprehensive measurement. To evaluate the performance of our method, we introduce a set of evaluation metrics and conduct extensive experiments. Results show that our method achieves good performance on both automatic and human evaluations.

pdf
CLGC: A Corpus for Chinese Literary Grace Evaluation
Yi Li | Dong Yu | Pengyuan Liu
Proceedings of the Thirteenth Language Resources and Evaluation Conference

In this paper, we construct a Chinese literary grace corpus, CLGC, with 10,000 texts and more than 1.85 million tokens. Multi-level annotations are provided for each text in our corpus, including literary grace level, sentence category, and figure-of-speech type. Based on the corpus, we dig deep into the correlation between fine-grained features (semantic information, part-of-speech and figure-of-speech, etc.) and literary grace level. We also propose a new Literary Grace Evaluation (LGE) task, which aims at making a comprehensive assessment of the literary grace level according to the text. In the end, we build some classification models with machine learning algorithms (such as SVM, TextCNN) to prove the effectiveness of our features and corpus for LGE. The results of our preliminary classification experiments have achieved 79.71% on the weighted average F1-score.

pdf bib
Analysis of Gender Bias in Social Perception and Judgement Using Chinese Word Embeddings
Jiali Li | Shucheng Zhu | Ying Liu | Pengyuan Liu
Proceedings of the 4th Workshop on Gender Bias in Natural Language Processing (GeBNLP)

Gender is a construction in line with social perception and judgment. An important means of this construction is through languages. When natural language processing tools, such as word embeddings, associate gender with the relevant categories of social perception and judgment, it is likely to cause bias and harm to those groups that do not conform to the mainstream social perception and judgment. Using 12,251 Chinese word embeddings as intermedium, this paper studies the relationship between social perception and judgment categories and gender. The results reveal that these grammatical gender-neutral Chinese word embeddings show a certain gender bias, which is consistent with the mainstream society’s perception and judgment of gender. Men are judged by their actions and perceived as bad, easily-disgusted, bad-tempered and rational roles while women are judged by their appearances and perceived as perfect, either happy or sad, and emotional roles.

2021

pdf
中文句子级性别无偏数据集构建及预训练语言模型的性别偏度评估(Construction of Chinese Sentence-Level Gender-Unbiased Data Set and Evaluation of Gender Bias in Pre-Training Language)
Jishun Zhao (赵继舜) | Bingjie Du (杜冰洁) | Shucheng Zhu (朱述承) | Pengyuan Liu (刘鹏远)
Proceedings of the 20th Chinese National Conference on Computational Linguistics

自然语言处理领域各项任务中,模型广泛存在性别偏见。然而当前尚无中文性别偏见评估和消偏的相关数据集,因此无法对中文自然语言处理模型中的性别偏见进行评估。首先本文根据16对性别称谓词,从一个平面媒体语料库中筛选出性别无偏的句子,构建了一个含有20000条语句的中文句子级性别无偏数据集SlguSet。随后,本文提出了一个可衡量预训练语言模型性别偏见程度的指标,并对5种流行的预训练语言模型中的性别偏见进行评估。结果表明,中文预训练语言模型中存在不同程度的性别偏见,该文所构建数据集能够很好的对中文预训练语言模型中的性别偏见进行评估。同时,该数据集还可作为评估预训练语言模型消偏方法的数据集。

pdf
中文关系抽取的句级语言学特征探究(A Probe into the Sentence-level Linguistic Features of Chinese Relation Extraction)
Baixi Xing (邢百西) | Jishun Zhao (赵继舜) | Pengyuan Liu (刘鹏远)
Proceedings of the 20th Chinese National Conference on Computational Linguistics

神经网络模型近些年在关系抽取任务上已经展示出了很好的效果,然而我们对于特征提取的过程所知甚少,而这也进一步限制了深度神经网络模型在关系抽取任务上的进一步发展。当前已有研究工作对英文关系抽取的语言学特征进行探究,并且得到了一些规律。然而由于中文与西方语言之间明显的差异性,其所探究到的规律与解释性不适用于中文关系抽取。本文首次对中文关系抽取神经网络进行探究,采用了四个角度共13种探究任务,其中包含中文特有的分词探究任务。在两个关系抽取数据集上进行了实验,探究了中文关系抽取模型进行特征提取的规律。

pdf
BLCUFIGHT at SemEval-2021 Task 10: Novel Unsupervised Frameworks For Source-Free Domain Adaptation
Weikang Wang | Yi Wu | Yixiang Liu | Pengyuan Liu
Proceedings of the 15th International Workshop on Semantic Evaluation (SemEval-2021)

Domain adaptation assumes that samples from source and target domains are freely accessible during a training phase. However, such assumption is rarely plausible in the real-world and may causes data-privacy issues, especially when the label of the source domain can be a sensitive attribute as an identifier. SemEval-2021 task 10 focuses on these issues. We participate in the task and propose novel frameworks based on self-training method. In our systems, two different frameworks are designed to solve text classification and sequence labeling. These approaches are tested to be effective which ranks the third among all system in subtask A, and ranks the first among all system in subtask B.

pdf
A Comparative Study of Collocation Extraction Methods from the Perspectives of Vocabulary and Grammar: A Case Study in the Field of Journalism
Lulu Gu | Yue Pan | Pengyuan Liu
Proceedings of the 35th Pacific Asia Conference on Language, Information and Computation

2020

pdf
Sensorimotor Enhanced Neural Network for Metaphor Detection
Mingyu Wan | Baixi Xing | Qi Su | Pengyuan Liu | Chu-Ren Huang
Proceedings of the 34th Pacific Asia Conference on Language, Information and Computation

pdf
Imbalanced Chinese Multi-label Text Classification Based on Alternating Attention
Hongliang Bi | Han Hu | Pengyuan Liu
Proceedings of the 34th Pacific Asia Conference on Language, Information and Computation

pdf bib
基于语料库的武侠与仙侠网络小说文体、词汇及主题对比分析(A Corpus-based Contrastive Analysis of Style, Vocabulary and Theme of Wuxia and Xianxia Internet Novels)
Sanle Zhang (张三乐) | Pengyuan Liu (刘鹏远) | Hu Zhang (张虎)
Proceedings of the 19th Chinese National Conference on Computational Linguistics

网络文学在我国发展迅猛,其数量和影响力呈现逐年上升的趋势,但目前尚无公开的较大规模网络文学作品语料库,鲜见基于语料库对网络文学具体类别作品的定量研究。本文初步建立了一个网络文学语料库,其中包括武侠和仙侠网络小说,使用文本计量、词频统计以及主题挖掘的方法对两类小说的文体风格、具体词汇使用和小说主题进行对比分析。通过比较,我们发现两类小说的文体风格大致相同,它们在词汇的使用和主题上既有共性又各具特色。从微观到宏观,从表面到内容,将定量统计和定性分析相结合,多角度、多层次的对武侠和仙侠网络小说进行比较。

pdf
基于计量的百年中国人名用字性别特征研究(A Quantified Research on Gender Characteristics of Chinese Names in A Century)
Bingjie Du (杜冰洁) | Pengyuan Liu (刘鹏远) | Yongsheng Tian (田永胜)
Proceedings of the 19th Chinese National Conference on Computational Linguistics

本文构建了一个包含11万以上条目规模的中国名人人名数据库,每条数据含有人名、性别、出生地等社会文化标签,同时含有拼音、笔画、偏旁等文字信息标签,这是目前已知最大的可用于研究的汉语真人人名数据库。基于该数据库,本文从中选择1919年至今的人名,用定性与定量结合的方法探究人名中汉字的特征和其性别差异以及历时变化。从人名长度来看,男性人名比女性人名长;从人名用字的难易度来看,女性用字比男性更复杂;从用字丰富度来看,人名用字越来越单一和集中化,男性人名的用字丰富度大于女性人名。计算人名用字的性别偏度后发现女性人名的专用自更多。两性用字意象有明显的不同,用字的意象随着时间发生改变,但改变最明显的时间节点是改革开放前后,其中女性的变化比男性显著。除此之外,我们还得出人名中的性别极性字表、各个阶段的高频字表、用字变化趋势表等。

pdf
伟大的男人和倔强的女人:基于语料库的形容词性别偏度历时研究(Great Males and Stubborn Females: A Diachronic Study of Corpus-Based Gendered Skewness in Chinese Adjectives)
Shucheng Zhu (朱述承) | Pengyuan Liu (刘鹏远)
Proceedings of the 19th Chinese National Conference on Computational Linguistics

性别偏见现象是社会语言学和计算语学学者均关注的研究热点,但目前大多数研究都是基于英语的,鲜有对汉语中性别偏见现象,特别是基于形容词的研究缺乏。而形容词是衡量社会对男性和女性角色规约的有力抓手。本文首先利用调查问卷的方法,构建了一个含有466个形容词的数据集,定义性别偏度为特定形容词词义和男性或女性群体相匹配的程度,并计算了数据集中每个形容词的性别偏度。然后基于DCC语料库,研究了《人民日报》的形容词性别偏度的历时总体变化,并考察了和姓名搭配的形容词的历时变化。发现《人民日报》所使用的形容词随时间的推移整体呈现中性化趋势,但在文化大革命期间呈现非常男性化的特征,和男性姓名搭配的形容词整体呈现中性化趋势。

pdf
小样本关系分类研究综述(Few-Shot Relation Classification: A Survey)
Han Hu (胡晗) | Pengyuan Liu (刘鹏远)
Proceedings of the 19th Chinese National Conference on Computational Linguistics

关系分类作为构建结构化知识的重要一环,在自然语言处理领域备受关注。但在很多应用领域中(医疗、金融领域),收集充足的用于训练关系分类模型的数据是十分困难的。近年来,仅需要少量训练样本的小样本学习研究逐渐新兴于各大领域。本文对近期小样本关系分类模型与方法进行了系统的综述。根据度量方法的不同,将现有方法分为原型式和分布式两大类。根据是否利用额外信息,将模型分为预训练和非预训练两大类。此外,除了常规设定下的小样本学习,本文还梳理了跨领域和稀缺资源场景下的小样本学习,并探讨了目前小样本关系分类方法的局限性,分析了跨领域小样本 学习面临的技术挑战。最后,展望了小样本关系分类未来的发展方向。

pdf
CDCPP:跨领域中文标点符号预测(CDCPP: Cross-Domain Chinese Punctuation Prediction)
Pengyuan Liu (刘鹏远) | Weikang Wang (王伟康) | Likun Qiu (邱立坤) | Bingjie Du (杜冰洁)
Proceedings of the 19th Chinese National Conference on Computational Linguistics

标点符号对文本理解起很大作用。但目前,在中文文本特别是在社交媒体及问答领域文本中的标点符号使用存在非常多的错误或缺失的情况,这严重影响对其进行语义分析及机器翻译等各项自然语言处理的效果。当前对标点符号进行预测的相关研究多集中于英文对话的语音转写文本,缺少对社交媒体及问答领域文本进行标点预测的相关研究,也没有这些领域公开的数据集。本文首先提出跨领域中文标点符号预测任务,该任务是要利用标点符号基本规范正确的大规模新闻领域文本,建立标点符号预测模型,然后在标点符号标注不规范的社交媒体及问答领域,进行跨领域标点符号预测。随后构建了新闻、社交媒体及问答三个领域的相应数据集。最后还实现了一个基于BERT的标点符号预测基线模型,并在该数据集上进行了实验与分析。实验结果表明,直接利用新闻领域训练的模型,在社交媒体及问答领域上进行标点符号预测的性能均有所下降,在问答领域下降较小,在微博领域下降较大,超过20%,跨领域标点符号预测任务具有一定的挑战性。

pdf
多目标情感分类中文数据集构建及分析研究(Construction and Analysis of Chinese Multi-Target Sentiment Classification Dataset)
Pengyuan Liu (刘鹏远) | Yongsheng Tian (田永胜) | Chengyu Du (杜成玉) | Likun Qiu (邱立坤)
Proceedings of the 19th Chinese National Conference on Computational Linguistics

目标级情感分类任务是要得到句子中特定评价目标的情感倾向。一个评论句中往往存在多个目标,多个目标的情感可能一致,也可能不一致。但在已有针对目标级情感分类的评测数据集中:1)大多数是一个句子一个目标;2)在少数有多个目标的句子中,多个目标情感倾向分布很不均衡,多个目标情感一致的情形占较大优势。数据集本身的缺陷限制了模型针对多个目标进行情感分类的提升空间。针对以上问题,本文构建了一个针对多目标情感分类的中文数据集,人工标注了6339个评价目标,共2071条数据。该数据集:1)评价目标个数分布平衡;2)情感正负极性分布平衡;3)多目标情感倾向分布平衡。随后,本文利用多个目标情感分类的主流模型在该数据集上进行了实验与比较分析。结果表明现有主流模型尚不能对存在多个目标且目标情感倾向性不一致实例中的目标进行很好的分类,尤其是目标的情感倾向为中性时。多目标情感分类任务具有一定的难度与挑战性。

2015

pdf
A Corpus-based Comparatively Study on the Semantic Features and Syntactic patterns of Yòu/Hái in Mandarin Chinese
Yuncui Zhang | Pengyuan Liu
Proceedings of the 29th Pacific Asia Conference on Language, Information and Computation: Posters

2010

pdf
PKU_HIT: An Event Detection System Based on Instances Expansion and Rich Syntactic Features
Shiqi Li | Pengyuan Liu | Tiejun Zhao | Qin Lu | Hanjing Li
Proceedings of the 5th International Workshop on Semantic Evaluation

pdf
PengYuan@PKU: Extracting Infrequent Sense Instance with the Same N-Gram Pattern for the SemEval-2010 Task 15
Peng-Yuan Liu | Shi-Wen Yu | Shui Liu | Tie-Jun Zhao
Proceedings of the 5th International Workshop on Semantic Evaluation

pdf
Combining Constituent and Dependency Syntactic Views for Chinese Semantic Role Labeling
Shiqi Li | Qin Lu | Tiejun Zhao | Pengyuan Liu | Hanjing Li
Coling 2010: Posters

pdf
Head-modifier Relation based Non-lexical Reordering Model for Phrase-Based Translation
Shui Liu | Sheng Li | Tiejun Zhao | Min Zhang | Pengyuan Liu
Coling 2010: Posters

2007

pdf
HIT-WSD: Using Search Engine for Multilingual Chinese-English Lexical Sample Task
PengYuan Liu | TieJun Zhao | MuYun Yang
Proceedings of the Fourth International Workshop on Semantic Evaluations (SemEval-2007)