文献类型：期刊文献

中文题名：A multi-target stance detection based on Bi-LSTM network with position-weight

英文题名：A multi-target stance detection based on Bi-LSTM network with position-weight

作者：Xu, Yilong[1]; Li, Wenfa[2]; Wang, Gongming[3,4]; Huang, Lingyun[5]

第一作者：Xu, Yilong

通讯作者：Li, Wenfa

机构：[1] Smart City College, Beijing Union University, Beijing, 100101, China; [2] College of Robotics, Beijing Union University, Beijing, 100101, China; [3] Tianyuan Network Co., Ltd., Beijing, 100193, China; [4] Beijing Tianyuan Network Co., Ltd., Beijing, 100193, China; [5] Chinatelecom Information Development Co., Ltd., Beijing, 100093, China

第一机构：北京联合大学智慧城市学院

通讯机构：[2]College of Robotics, Beijing Union University, Beijing, 100101, China|[1141739]北京联合大学机器人学院;[11417]北京联合大学;

年份：2020

卷号：26

期号：4

起止页码：442-447

中文期刊名：高技术通讯：英文版

外文期刊名：High Technology Letters

收录：EI(收录号：20205209671461);Scopus(收录号：2-s2.0-85097906576)

基金：Supported by the National Natural Science Foundation of China(No.61972040);the Science and Technology Projects of Beijing Municipal Education Commission(No.KM201711417011);the Premium Funding Project for Academic Human Resources Development in Beijing Union University(No.BPHR2020AZ03)。

语种：英文

中文关键词：long short-term memory(LSTM);multi-target;natural language processing;stance detection

外文关键词：Computer science

摘要：In the task of multi-target stance detection,there are problems the mutual influence of content describing different targets,resulting in reduction in accuracy.To solve this problem,a multi-target stance detection algorithm based on a bidirectional long short-term memory(Bi-LSTM)network with position-weight is proposed.First,the corresponding position of the target in the input text is calculated with the ultimate position-weight vector.Next,the position information and output from the Bi-LSTM layer are fused by the position-weight fusion layer.Finally,the stances of different targets are predicted using the LSTM network and softmax classification.The multi-target stance detection corpus of the American election in 2016 is used to validate the proposed method.The results demonstrate that the Bi-LSTM network with position-weight achieves an advantage of 1.4%in macro average F1 value in the comparison of recent algorithms.
In the task of multi-target stance detection, there are problems the mutual influence of content describing different targets, resulting in reduction in accuracy. To solve this problem, a multi-target stance detection algorithm based on a bidirectional long short-term memory (Bi-LSTM) network with position-weight is proposed. First, the corresponding position of the target in the input text is calculated with the ultimate position-weight vector. Next, the position information and output from the Bi-LSTM layer are fused by the position-weight fusion layer. Finally, the stances of different targets are predicted using the LSTM network and softmax classification. The multi-target stance detection corpus of the American election in 2016 is used to validate the proposed method. The results demonstrate that the Bi-LSTM network with position-weight achieves an advantage of 1.4% in macro average F1 value in the comparison of recent algorithms. Copyright ? by HIGH TECHNOLOGY LETTERS PRESS.