摘要
受中美贸易摩擦不断升级、国内经济结构调整和金融市场波动等环境压力影响,目前全国就业形势整体较为严峻。互联网中海量岗位信息的存在,为求职者和招聘单位带来了便捷,也为精准人岗匹配提出了挑战。基于内容的推荐算法较适用于人岗匹配,但是目前大多数方法使用的特征较少,特别是对提供的长文本信息利用不够。本文提出一种结合深度语义特征的人岗精准匹配算法,在构建较为完善的人岗特征体系基础上,利用自然语言处理技术,采用Doc2vec方法充分挖掘长文本中包含的语义信息,实现求职者与岗位之间信息的精准匹配。该方法既能克服数据稀疏和冷启动问题,同时能充分利用求职者和岗位提供的信息,有利于实现更加准确、个性化的就业推荐服务。
Influenced by the escalating trade friction between China and the United States,the adjustment of domestic economic structure and the fluctuation of financial market,the employment situation in China is more severe as a whole. The existence of a large amount of job information on the Internet has brought convenience to job seekers and recruiters,as well as challenges to accurate job matching. Content-based recommendation algorithm is more suitable for job-to-post matching,but at present most methods use fewer features,especially the use of the long text information provided is not enough.. This paper proposes an accurate human-post matching algorithm combined with deep semantic features. On the basis of building a relatively complete human-post feature system,natural language processing technology and Doc2 vec method are used to fully mine the semantic information contained in long texts,so as to realize the accurate matching of information between job seekers and positions. This method can not only overcome the problem of data sparseness and cold start,but also make full use of the information provided by job seekers and positions,which is conducive to achieving more accurate and personalized employment recommendation services.
作者
张毅
高元荣
黄宗财
吴升
王毅青
黄幼姑
ZHANG Yi;GAO Yuanrong;HUANG Zongcai;WU Sheng;WANG Yiqing;HUANG Yougu(Fujian Star Big Data Application Service Co.,Ltd.,Fuzhou 350003,China;Digital China Research Institute,Fuzhou University,Fuzhou 350002,China)
出处
《贵州大学学报（自然科学版）》
2021年第1期65-70,共6页
Journal of Guizhou University:Natural Sciences
基金
星空大数据应用技术联合实验室开放基金资助项目(NBD-2018-165)。
