结合单词-字符引导注意力网络的中文旅游文本命名实体识别

来源 :计算机工程 | 被引量 : 0次 | 上传用户:and113
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
传统基于词向量表示的命名实体识别方法通常忽略了字符语义信息、字符间的位置信息,以及字符和单词间的关联关系.提出一种基于单词-字符引导注意力网络(WCGAN)的中文旅游命名实体识别方法,利用单词引导注意力网络获取单词间的序列信息和关键单词信息,采用字符引导注意力网络捕获字符语义信息和字符间的位置信息,增强单词和字符间的关联性与互补性,从而实现中文旅游文本中命名实体的识别.实验结果表明,WCGAN方法在ResumeNER和 TourismNER基准数据集上的F值分别为93.491% 和92.860%,相比 Bi-LSTM+CRF、Char-Dense等方法识别效果更好.
其他文献
Evaluation is an essential part of the teaching process, especially in the programming course. Both students and teachers can benefit significantly from automat
To build the top class of higher education by focusing on the goal of the “Double First-Class”, it needs to analyze the defect and deficiency and identify the
With more and more colleges and universities set up artificial intelligence undergraduate major, the cultivation of artificial intelligence undergraduate has be
It is a challenge to cultivate senior and comprehensive talents in software engineering. To promote the quality of software engineering education, this paper pr