| 采用多尺度注意力机制的远程监督关系抽取 |
| |
| 引用本文： | 蔡强,郝佳云,曹健,李海生.采用多尺度注意力机制的远程监督关系抽取[J].中文信息学报,2018,32(1):96-101. |
| |
| 作者姓名： | 蔡强 郝佳云 曹健 李海生 |
| |
| 作者单位： | 1.北京工商大学 计算机与信息工程学院,北京 100048;2.北京工商大学 食品安全大数据技术北京市重点实验室,北京 100048 |
| |
| 基金项目： | 北京市教委科研计划一般项目(SQKM201610011010);北京市自然科学基金(4162019);北京市科技计划课题(Z161100001616004) |
| |
| 摘 要： | 针对目前大多数关系抽取模型中局部特征及全局特征利用不充分的缺点,该文提出一种采用多尺度注意力机制的远程监督关系抽取模型。在词语层面,通过在池化层构建权重矩阵来衡量词语与关系的相关程度,从而捕捉句子中重要的语义特征;在句子层面,采用注意力机制将预测关系与句子进行相关性比较,获得句子级别的重要信息。模型在NYT数据集上平均准确率达到78%,表明该模型能够有效地利用多尺度特征,并且提高远程关系抽取任务的准确率。
|
| 关 键 词： | 多尺度 注意力机制 远程监督模型 关系抽取 |
Multi-level Attention Mechanism Based Distant Supervision for Relation Extraction |
| |
| CAI Qiang,HAO Jiayun,CAO Jian,LI Haisheng.Multi-level Attention Mechanism Based Distant Supervision for Relation Extraction[J].Journal of Chinese Information Processing,2018,32(1):96-101. |
| |
| Authors: | CAI Qiang HAO Jiayun CAO Jian LI Haisheng |
| |
| Affiliation: | 1. School of Computer and Information Engineering, Beijing Technology and Business University, Beijing 100048, China; 2. Beijing Key Laboratory of Big Data Technology for Food Safety, Beijing Technology and Business University, Beijing 100048, China |
| |
| Abstract: | To best exploit the local and global feature, we proposed a distant supervised relation extraction model based on multi-level attention mechanism. We employ an attention matrix in pooling layer to capture the word-level sematic feature which indicates the relevant relationship between input words and relations. Moreover, we adopted sentence-level attention mechanism to compare the relationship between sentences and predicted relations. Experimental results show that the mean accuracy of the proposed model achieves 78% in the NYT data set, indicating an effective use of multi-level feature and better performance of distant relation extraction task. |
| |
| Keywords: | multi-level attention mechanism distant supervision model relation extraction |
|
| 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 |
|
