图书情报工作 ›› 2022, Vol. 66 ›› Issue (14): 101-118.DOI: 10.13266/j.issn.0252-3116.2022.14.011

• 情报研究 • 上一篇    下一篇

基于招聘文本实体挖掘的人才供需分析——以人工智能领域为例

袁毅, 陶鑫琪, 李瑾萱, 刘娅娴, 汪晓芸, 景香玉   

  1. 华东师范大学经济与管理学部信息管理系 上海 200062
  • 收稿日期:2022-03-01 修回日期:2022-05-09 出版日期:2022-07-20 发布日期:2022-07-28
  • 作者简介:袁毅,教授,博士,E-mail:yuanyixz@126.com;陶鑫琪,硕士研究生;李瑾萱,硕士研究生;刘娅娴,硕士研究生;汪晓芸,硕士研究生;景香玉,硕士研究生。
  • 基金资助:
    本文系上海市2020年度"科技创新行动计划"软科学重点项目"人工智能复合型人才需求及培养模式研究"(项目编号:20692108300)研究成果之一。

Analysis of Talent Supply and Demand Based on Recruitment Text Entity Mining——Taking Artificial Intelligence Field as an Example

Yuan Yi, Tao Xinqi, Li Jinxuan, Liu Yaxian, Wang Xiaoyun, Jing Xiangyu   

  1. Department of Information Management, Faculty of Economics and Management, East China Normal University, Shanghai 200062
  • Received:2022-03-01 Revised:2022-05-09 Online:2022-07-20 Published:2022-07-28

摘要: [目的/意义]基于网络招聘文本和学科数据,提出"行业-岗位-知识-学科"的人才需求及供给分析框架,以人工智能领域为例进行挖掘与分析,同时对其他领域的人才供需分析也具有借鉴意义。[方法/过程]采集招聘网站中与人工智能相关的职位招聘公告,综合对比CRF、BiLSTM-CRF、BERT-BiLSTM-CRF、BERT模型对招聘文本的实体抽取效果,并运用社会网络分析方法与学科数据进行关联分析。[结果/结论]BERT-BiLSTM-CRF实体抽取实验效果最佳,分别构建"行业-岗位""岗位-知识"以及"知识-学科"3种关系网络,得到与人工智能领域联系最紧密的行业、岗位、知识及学科。该框架能充分地挖掘人才需求现状,并能较精准地将需求定位到人才培养的学科,对于国家发展战略以及高等院校人才培养计划的制订具有现实意义。

关键词: 招聘实体, 人工智能, 供需分析, 人才培养, 深度学习

Abstract: [Purpose/significance] Based on online recruitment texts and subject data, this paper proposes an analysis framework of talent demand and supply of "Industry-Position-Knowledge-Discipline", takes the artificial intelligence industry as an example for mining and analysis, and it also has references and significance for the analysis of supply and demand of talents in other disciplines. [Method/process] This paper collected job announcements related to artificial intelligence from recruitment websites, and compared the effect of CRF, BiLSTM-CRF, BERT-BiLSTM-CRF, and BERT models on entity extraction of recruitment texts, and used social network analysis methods to conduct correlation analysis with subject data. [Result/conclusion] The BERT-BiLSTM-CRF entity extraction experiment is the best. Three relational networks of "industry-post", "post-knowledge" and "knowledge-discipline" are constructed respectively, and the industry, positions, knowledge and disciplines most closely related to artificial intelligence disciplines are obtained. The framework can fully explore the current situation of talent demand and locate the demand to the disciplines of talent cultivation more precisely, which is of practical significance for the national development strategy and the formulation of talent cultivation plan of higher education institutions.

Key words: recruitment entity, artificial intelligence, supply and demand analysis, talent training, deep learning

中图分类号: