no code implementations • CCL 2020 • Xiaodong Yan, Xiaoqing Xie
句子排序是多文档摘要系统和机器阅读理解中重要的任务之一, 排序的质量将直接 影响摘要和答案的连贯性与可读性。因此, 本文采用在中英文上大规模使用的深度 学习方法, 同时结合朝鲜语词语形态变化丰富的特点, 提出了一种基于子词级别词 向量和指针网络的朝鲜语句子排序模型, 其目的是解决传统方法无法挖掘深层语义 信息问题。 本文提出基于形态素拆分的词向量训练方法(MorV), 同时对比子词n元 词向量训练方法(SG), 得到朝鲜语词向量;采用了两种句向量方法:基于卷积神经网 络(CNN)、基于长短时记忆网络(LSTM), 结合指针网络分别进行实验。结果表明本文 采用MorV和LSTM的句向量结合方法可以更好地捕获句子间的语义逻辑关系, 提升句 子排序的效果。 关键词: 词向量 ;形态素拆分 ;指针网络 ;句子排序
no code implementations • CCL 2020 • Xiaodong Yan, Xiaoqing Xie, Yu Zou, Wei Li
Seq2seq神经网络模型在中英文文本摘要的研究中取得了良好的效果, 但在低资源语言的文本摘要研究还处于探索阶段, 尤其是在藏语中。此外, 目前还没有大规模的标注语料库进行摘要提取。本文提出了一种生成藏文新闻摘要的统一模型。利用TextRank算法解决了藏语标注训练数据不足的问题。然后, 采用两层双GRU神经网络提取代表原始新闻的句子, 减少冗余信息。最后, 使用基于注意力机制的Seq2Seq来生成理解式摘要。同时, 我们加入了指针网络来处理未登录词的问题。实验结果表明, ROUGE-1评分比传统模型提高了2%。 关键词:文本摘要;藏文;TextRank; 指针网络;Bi-GRU
1 code implementation • 10 Feb 2023 • Lei Zhang, Xiaodong Yan, Jianshan He, Ruopeng Li, Wei Chu
Our experimental results show that our model effectively relieves the problem of over-smoothing in deep GCNs and outperforms the state-of-the-art (SOTA) methods on various benchmark datasets.
no code implementations • 12 Nov 2022 • Zengjing Chen, Huaijin Liang, Wei Wang, Xiaodong Yan
No matter how much some gamblers occasionally win, as long as they continue to gamble, sooner or later they will lose more to the casino, which is the so-called long bet will lose.
no code implementations • 20 Oct 2021 • Wei Wang, Xiaodong Yan, Yanyan Ren, Zhijie Xiao
Heterogeneous panel data models that allow the coefficients to vary across individuals and/or change over time have received increasingly more attention in statistics and econometrics.
no code implementations • 7 Oct 2021 • Ke Sun, Yingnan Zhao, Enze Shi, Yafei Wang, Xiaodong Yan, Bei Jiang, Linglong Kong
The theoretical advantages of distributional reinforcement learning~(RL) over classical RL remain elusive despite its remarkable empirical performance.
no code implementations • 29 Sep 2021 • Ke Sun, Yingnan Zhao, Yi Liu, Enze Shi, Yafei Wang, Aref Sadeghi, Xiaodong Yan, Bei Jiang, Linglong Kong
Distributional reinforcement learning~(RL) is a class of state-of-the-art algorithms that estimate the whole distribution of the total return rather than only its expectation.