文献类型：期刊文献

英文题名：A comprehensive exploration of semantic relation extraction via pre-trained CNNs

作者：Li, Qing[1];Li, Lili[2];Wang, Weinan[3];Li, Qi[4];Zhong, Jiang[1,5]

机构：[1]Chongqing Univ, Coll Comp Sci, Chongqing, Peoples R China;[2]Chongqing Univ, Sch Civil Engn, Chongqing, Peoples R China;[3]Peking Univ, Sch Math Sci, Beijing, Peoples R China;[4]Shaoxing Univ, Dept Comp Sci & Engn, Shaoxing, Peoples R China;[5]Chongqing Univ, Key Lab Dependable Serv Comp Cyber Phys Soc, Chongqing, Peoples R China

年份：2020

卷号：194

外文期刊名：KNOWLEDGE-BASED SYSTEMS

收录：SCI-EXPANDED(收录号：WOS:000534581300006)、、EI(收录号：20200508117981)、Scopus(收录号：2-s2.0-85078739521)、WOS

基金：We are grateful to the anonymous reviewers for their valuable comments on this manuscript. The research has been supported by the National Key Research and Development Program of China Grant 2017YFB1402401, in part by the Fundamental Research Funds for the Central Universities under Grant 2018CDYJSY0055, in part by the Graduate Research and Innovation Foundation of Chongqing under Grant CYB18058, in part by the Social Undertakings and Livelihood Security Science and Technology Innovation Funds of CQ CSTC Grant cstc2017shmsA0641, in part by the Key Research Program of Chongqing Science and Technology Bureau under Grant cstc2018jszx-cyzdX0086, cstc2019jscx-fxyd0142 and cstc2017zdcy-zdyf0150.

语种：英文

外文关键词：Relation extraction; Semantic relation; Natural language processing; Convolutional neural networks

外文摘要：Semantic relation extraction between entity pairs is a crucial task in information extraction from text. In this paper, we propose a new pre-trained network architecture for this task, and it is called the XM-CNN. The XM-CNN utilizes word embedding and position embedding information. It is designed to reinforce the contextual output from the MT-DNNKD pre-trained model. Our model effectively utilized an entity-aware attention mechanisms to detected the features and also adopts and applies more relation-specific pooling attention mechanisms applied to it. The experimental results show that the XM-CNN achieves state-of-the-art results on the SemEval-2010 task 8, and a thorough evaluation of the method is conducted. (C) 2020 Elsevier B.V. All rights reserved.