计算机与数字工程2024，Vol.52Issue(1)：140-144,5.DOI:10.3969/j.issn.1672-9722.2024.01.022
基于词向量和卷积神经网络的税务问答系统
Tax Q&A System Based on Word Vector and Convolutional Neural Network
杜航 1牟莉1
作者信息
- 1. 西安工程大学计算机科学学院 西安 710600
- 折叠
摘要
Abstract
Tax consultation Q&A business volume is large,requires a high degree of professionalism,and in recent years,the new tax preferential policies are more and more,taxpayers'demand for tax related laws and regulations is also growing.Most of the existing ways of tax consultation are on-site consultation or telephone consultation in the tax hall.Tax workers need to constantly study and consult relevant policies and regulations for the problems of taxpayers,which makes the tax authorities face huge pressure of consultation.Nowadays,although some tax authorities are trying to use the Internet for online tax consultation,most of them are message consultation,which requires manual answers to the problems.A large number of hot issues are concentrated,causing sys-tem congestion,and are not instant,which can not effectively solve the current problems.Intelligent question answering technology can effectively solve the above problems.The tax question answering system based on this technology has the advantages of high effi-ciency and strong practicability,which provides more convenience for taxpayers and tax workers.关键词
税务咨询/智能问答/自然语言理解/神经网络Key words
tax consultation/intelligent question and answer/natural language processing comprehension/neural network分类
信息技术与安全科学引用本文复制引用
杜航,牟莉..基于词向量和卷积神经网络的税务问答系统[J].计算机与数字工程,2024,52(1):140-144,5.
