Improving entity linking with two adaptive features

作     者：Hongbin ZHANG Quan CHEN Weiwen ZHANG Hongbin ZHANG;Quan CHEN;Weiwen ZHANG

作者机构：School of Computer Science and TechnologyGuangdong University of TechnologyGuangzhou 510006China 

出 版 物：《Frontiers of Information Technology & Electronic Engineering》 (信息与电子工程前沿（英文版）)

年 卷 期：2022年第23卷第11期

页      面：1620-1630页

核心收录：

学科分类：0808[工学-电气工程] 081203[工学-计算机应用技术] 08[工学] 0835[工学-软件工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：Project supported by the Key-Area Research and Development Program of Guangdong Province,China(No.2019B010153002) the Program of Marine Economy Development(Six Marine Industries)Special Foundation of Department of Natural Resources of Guangdong Province,China(No.GDNRC056) the National Natural Science Foundation of China(No.62002071) the Top Youth Talent Project of Zhujiang Talent Program,China(No.2019QN01X516) the Guangdong Provincial Key Laboratory of Cyber-Physical System,China(No.2020B1212060069) 

主　　题：Entity linking Local model Global model Adaptive features Entity type 

摘      要：Entity linking(EL)is a fundamental task in natural language *** on neural networks,existing systems pay more attention to the construction of the global model,but ignore latent semantic information in the local model and the acquisition of effective entity type *** this paper,we propose two adaptive features,in which the first adaptive feature enables the local and global models to capture latent information,and the second adaptive feature describes effective information for entity type *** adaptive features can work together naturally to handle some uncertain entity type information for *** results demonstrate that our EL system achieves the best performance on the AIDA-B and MSNBC datasets,and the best average performance on out-domain *** results indicate that the proposed adaptive features,which are based on their own diverse contexts,can capture information that is conducive for EL.