基于有向异构图的发票明细税收分类方法(Tax Classification of Invoice Details Based on Directed Heterogeneous Graph)

Peiyao Zhao (赵珮瑶), Qinghua Zheng (郑庆华), Bo Dong (董博), Jianfei Ruan (阮建飞), Minnan Luo (罗敏楠)


Abstract
税收是国家赖以生存的物质基础。为加快税收现代化,方便纳税人便捷、规范开具增值税发票,国税总局规定纳税人在税控系统开票前选择发票明细对应的税收分类才可正常开具发票。提高税收分类的准确度,是构建税收风险指标和分析纳税人行为特征的重要基础。基于此,本文提出了一种基于有向异构图的短文本分类模型(Heterogeneous Directed Graph Attenton Network,HDGAT),利用发票明细间的有向信息建模,引入外部知识,显著地提高了发票明细的税收分类准确度。
Anthology ID:
2020.ccl-1.72
Volume:
Proceedings of the 19th Chinese National Conference on Computational Linguistics
Month:
October
Year:
2020
Address:
Haikou, China
Venue:
CCL
SIG:
Publisher:
Chinese Information Processing Society of China
Note:
Pages:
771–782
Language:
Chinese
URL:
https://aclanthology.org/2020.ccl-1.72
DOI:
Bibkey:
Cite (ACL):
Peiyao Zhao, Qinghua Zheng, Bo Dong, Jianfei Ruan, and Minnan Luo. 2020. 基于有向异构图的发票明细税收分类方法(Tax Classification of Invoice Details Based on Directed Heterogeneous Graph). In Proceedings of the 19th Chinese National Conference on Computational Linguistics, pages 771–782, Haikou, China. Chinese Information Processing Society of China.
Cite (Informal):
基于有向异构图的发票明细税收分类方法(Tax Classification of Invoice Details Based on Directed Heterogeneous Graph) (Zhao et al., CCL 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.ccl-1.72.pdf