no code implementations • CCL 2021 • Liang Liu, Fang Kong
“实体关系抽取旨在从文本中抽取出实体之间的语义关系, 是自然语言处理的一项基本任务。在新闻报道、维基百科等规范文本上该任务的研究相对丰富, 已经取得了一定的效果, 但面向对话文本的相关研究还处于起始阶段。相较于规范文本, 用于实体关系抽取的对话语料规模较小, 对话文本的有效特征难以捕获, 这使得面向对话文本的实体关系抽取更具挑战。该文针对这一任务提出了一个基于Star-Transformer的实体关系抽取模型, 通过融入高速网络进行信息桥接, 并在此基础上融入交互信息和知识, 最后使用多任务学习机制进一步提升模型的性能。在DialogRE公开数据集上实验得到F1值为55. 7%, F1c值为52. 3%, 证明了提出方法的有效性。”
1 code implementation • COLING 2022 • Yaxin Fan, Peifeng Li, Fang Kong, Qiaoming Zhu
Conversational discourse parsing aims to construct an implicit utterance dependency tree to reflect the turn-taking in a multi-party conversation.
no code implementations • CCL 2020 • Chun Chen, Mingyang Li, Fang Kong
中文社交媒体命名实体识别由于其领域特殊性, 一直广受关注。非正式且无结构的微博文本存在以下两个问题:一是词语边界模糊;二是语料规模有限。针对问题一, 本文将同维度的字词进行融合, 获得丰富的文本序列表征;针对问题二, 提出了基于Star-Transformer框架的命名实体识别模型, 借助星型拓扑结构更好地捕获动态特征;同时利用高速网络优化Star-Transformer中的信息桥接, 提升模型的鲁棒性。本文提出的轻量级命名实体识别模型取得了目前Weibo语料上最好的效果。
no code implementations • CCL 2020 • Bowen Si, Fang Kong
对话是一个顺序交互的过程, 回应选择旨在根据已有对话上文选择合适的回应, 是自然语言处理领域的研究热点。已有研究取得了一定的成功, 但仍然存在两个突出的问题。一是现有的编码器在挖掘对话文本语义信息上尚存在不足;二是只考虑每一回合对话与备选回应之间的关系, 忽视了对话上文的整体语义信息。针对问题一, 本文借助多头自注意力机制有效捕捉对话文本的语义信息;针对问题二, 整合对话上文的整体语义信息, 分别从单词、句子以及整体对话上文三个层次与备选回应进行匹配, 充分保证匹配信息的完整。在Ubuntu Corpus V1和Douban Conversation Corpus数据集上的对比实验表明了本文给出方法的有效性。
no code implementations • CCL 2021 • Qianying Dai, Fang Kong
“时序关系识别是信息抽取领域的一个重要分支, 对文本理解发挥着关键作用。按照关联对象的不同, 时序关系分为三大类:事件对(E-E)间的时序关系, 事件与时间表达式间(E-T)的时序关系, 事件与文档建立时间(E-D)间的时序关系。不同关系类型孤立识别的方法忽视了其间隐含的关联信息, 针对这一问题构建了基于信息交互增强的时序关系联合识别模型。通过在不同神经网络层之间共享参数实现E-E与E-T时序关系的语义交流, 利用两者的潜在联系提高识别精度。在Time-Bank Dense语料上的一系列实验表明, 该方法优于现有的大多数神经网络方法。”
1 code implementation • Findings (EMNLP) 2021 • Longyin Zhang, Xin Tan, Fang Kong, Guodong Zhou
Discourse analysis has long been known to be fundamental in natural language processing.
no code implementations • 11 Mar 2024 • Yu Xia, Fang Kong, Tong Yu, Liya Guo, Ryan A. Rossi, Sungchul Kim, Shuai Li
In this paper, we propose a time-increasing bandit algorithm TI-UCB, which effectively predicts the increase of model performances due to finetuning and efficiently balances exploration and exploitation in model selection.
no code implementations • 3 Jan 2024 • Fang Kong, Shuai Li
An extension of \citet{kong2023player} to this more general setting achieves a near-optimal bound for player-optimal regret.
no code implementations • 20 Jul 2023 • Fang Kong, Shuai Li
Most previous works in this line are only able to derive theoretical guarantees for player-pessimal stable regret, which is defined compared with the players' least-preferred stable matching.
1 code implementation • 19 May 2023 • Fang Kong, Jize Xie, Baoxiang Wang, Tao Yao, Shuai Li
The effect is neglected by previous OIM works under IC and linear threshold models.
no code implementations • 13 Mar 2023 • Fang Kong, Canzhe Zhao, Shuai Li
Follow-the-regularized-leader (FTRL) is another type of popular algorithm that can adapt to different environments.
no code implementations • 14 Feb 2023 • Fang Kong, Xiangcheng Zhang, Baoxiang Wang, Shuai Li
Learning Markov decision processes (MDP) in an adversarial environment has been a challenging problem.
no code implementations • 16 Jun 2022 • Fang Kong, Yichi Zhou, Shuai Li
With a general feedback graph, the observation of an arm may not be available when this arm is pulled, which makes the exploration more expensive and the algorithms more challenging to perform optimally in both environments.
1 code implementation • 26 Apr 2022 • Fang Kong, Junming Yin, Shuai Li
The problem of two-sided matching markets has a wide range of real-world applications and has been extensively studied in the literature.
no code implementations • NeurIPS 2021 • Fang Kong, Yueran Yang, Wei Chen, Shuai Li
These are the first theoretical results for TS to solve CMAB with a common approximation oracle and break the misconception that TS cannot work with approximation oracles.
1 code implementation • ACL 2021 • Longyin Zhang, Fang Kong, Guodong Zhou
Text-level discourse rhetorical structure (DRS) parsing is known to be challenging due to the notorious lack of training data.
1 code implementation • ACL 2021 • Chun Chen, Fang Kong
In comparison with English, due to the lack of explicit word boundary and tenses information, Chinese Named Entity Recognition (NER) is much more challenging.
no code implementations • COLING 2020 • Feng Jiang, Xiaomin Chu, Peifeng Li, Fang Kong, Qiaoming Zhu
Discourse structure tree construction is the fundamental task of discourse parsing and most previous work focused on English.
no code implementations • NeurIPS 2020 • Shuai Li, Fang Kong, Kejie Tang, Qizhi Li, Wei Chen
Based on the linear structure in node activations, we incorporate ideas from linear bandits and design an algorithm LT-LinUCB that is consistent with the observed feedback.
1 code implementation • ACL 2020 • Longyin Zhang, Yuqing Xing, Fang Kong, Peifeng Li, Guodong Zhou
Due to its great importance in deep natural language understanding and various down-stream applications, text-level parsing of discourse rhetorical structure (DRS) has been drawing more and more attention in recent years.
no code implementations • ACL 2019 • Sheng Xu, Peifeng Li, Fang Kong, Qiaoming Zhu, Guodong Zhou
In the literature, most of the previous studies on English implicit discourse relation recognition only use sentence-level representations, which cannot provide enough semantic information in Chinese due to its unique paratactic characteristics.