基于图卷积网络的岗位匹配模型研究
- 分享到:
摘要：
提出了一种基于图卷积网络(GCN)的岗位匹配模型，该模型首先对公司招聘信息和用户简历信息进行表示，进而利用图结构对用户简历和招聘需求信息进行建模，通过图卷积网络的传播与聚合操作，得到更能反映匹配关系的表示，最终实现招聘信息与简历信息的匹配。实验结果表明，所提的模型和基准模型相比，在F1值上提升了 3.3百分点，能有效提高招聘信息与用户简历信息的匹配性能。
We propose a job matching model based on Graph Convolutional Networks(GCN). The model firstly represents company job postings and user resumes, then models user resumes and job requirements using the graph structure in GCN, By leveraging graph convolutional network propagation and aggregation operations, a representation that better reflects the matching relationship is obtained, ultimately achieving the matching of job postings and resumes, Fxperimental results show that the model proposed in this paper improves the F1 score by 3.3 percentage points compared with the baseline model, effectively enhancing the matching performance between job postings and user resumes.
作者:
孙庆英，周含，刘思妍，李家宏，席倩
Sun Qingying, Zhou Han, Liu Siyan, Li Jiahong, Xi Qian,
机构地区：
淮阴师范学院计算机科学与技术学院
引用本文：
孙庆英，周含，刘思妍等。基于图卷积网络的岗位匹配模型研究[J].学报(自然科学版),2025,53(6) :93-100.(Sun Qingying,Zhou Han,Liu Siyan,et al.Job matching model based on graph convolutional networks[J].Journal of Henan Normal University(Natural Science Edition),2025,53(6):93-100.DOI:10.16366/j.cnki.1000-2367.2025.02.20.0002.)
基金：
国家自然科学基金；江苏高校“青蓝工程”优秀青年骨干教师资助项目；教育部人文社会科学研究一般项目；江苏省大数据智能工程实验室开放课题
关键词：
图卷积网络；图结构；岗位匹配；自然语言处理
graph convolutional networks; graph structure; job matching; natural language processing
分类号：
TP391
