|
|||||
|
|
| 一种针对成分树的混合神经网络模型 | |
| 引用本文： | 霍欢,薛瑶环,黄君扬,金轩城,邹依婷.一种针对成分树的混合神经网络模型[J].中文信息学报,2019,33(3):8-16. |
| 作者姓名： | 霍欢 薛瑶环 黄君扬 金轩城 邹依婷 |
| 作者单位： | 1.上海理工大学 光电信息与计算机工程学院,上海 200093; 2.复旦大学 上海市数据科学重点实验室,上海 201203 |
| 基金项目： | 国家自然科学基金(61003031);上海重点科技攻关项目(14511107902);上海市工程中心建设项目(GCZX14014);上海市一流学科建设项目(XTKX2012);上海市数据科学重点实验室开放课题(201609060003);沪江基金研究基地专项(C14001) |
| 摘 要： | 为了提高自然语言处理的准确度,很多工作将句法成分树与LSTM相结合,提出了各种针对成分树的LSTM模型(文中用C-TreeLSTM统称这类模型)。考虑到C-TreeLSTM模型在计算内部节点隐藏状态的过程中,由于一个重要信息来源(即单词)的缺失导致文本建模的准确度不高,该文提出一种针对成分树的混合神经网络模型,通过在C-TreeLSTM模型的节点编码过程中注入各节点所覆盖的短语语义向量来增强节点对文本语义的记忆,故将此模型命名为SC-TreeLSTM。实验结果表明,该模型在情感分类和机器阅读理解两类任务上表现优异。 |
| 关 键 词： | 成分树 C-TreeLSTM 短语语义向量 混合模型 |
A Hybrid Neural Network Model on Constituent Tree Structure |
|
| HUO Huan,XUE Yaohuan,HUANG Junyang,JIN Xuancheng,ZOU Yiting.A Hybrid Neural Network Model on Constituent Tree Structure[J].Journal of Chinese Information Processing,2019,33(3):8-16. | |
| Authors: | HUO Huan XUE Yaohuan HUANG Junyang JIN Xuancheng ZOU Yiting |
| Affiliation: | 1.School of Optical-electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China; 2.Shanghai Key Laboratory of Data Science, Fudan University, Shanghai 201203, China |
| Abstract: | Current methods of combining constituent trees with LSTM (C-TreeLSTM) suffere from low accuracy for text modeling due to withouth computing the words in hidden state of internal nodes. This paper proposes a hybrid neural network model, i.e. SC-TreeLSTM, based on the constituent tree structure. The model enhances nodes memory of text semantics by injecting phrase semantic vectors which is covered by corresponding node during encoding. The experimental results show that the SC-TreeLSTM achieves excellent performance in both sentiment classification and machine reading comprehension tasks. |
| Keywords: | constituent tree C-TreeLSTM phrase semantic vector hybrid model |
| 本文献已被 维普 等数据库收录！ | |
| 点击此处可从《中文信息学报》浏览原始摘要信息 | |
| 点击此处可从《中文信息学报》下载全文 | |