基于混合神经网络的开源社区软件开发者人力资源价值预测

HUMAN RESOURCE VALUE PREDICTION OF OPEN SOURCE COMMUNITY SOFTWARE DEVELOPERS BASED ON HYBRID NEURAL NETWORK

随着市场经济的不断发展,人力资源也逐渐受到社会各界的关注。在IT行业中,由于技术的快速发展与更新换代,企业内对高新技术人才的需求日益加大,企业间对稀缺人才的争夺呈现白热化状态。现有的管理学评估方法难以满足企业需求,为更高效地发掘GitHub中的软件开发人才,基于GitHub软件开发者编程能力、项目管理能力、学习能力、团队合作能力、技术影响力、敬业度建立软件开发者人力资源价值评估体系。使用CNN-LSTM混合神经网络进行软件开发者人力资源价值评估并预测未来价值。实验表明,模型评估软件开发者价值的准确率可达98.59%,测试集上软件开发者人力资源价值预测结果符合企业招聘者预估价值。

With the continuous development of market economy,human resources are gradually concerned by all walks of life.In the IT industry,due to the rapid development and upgrading of technology,the demand for high-tech talents in enterprises is increasing day by day,and the competition for scarce talents among enterprises is becoming increasingly fierce.However,the existing economic evaluation methods can not meet the needs of enterprises.In order to explore the software development talents in GitHub more efficiently,this paper establishes a software developer human resource value evaluation system based on GitHub software developer s programming ability,project management ability,learning ability,team cooperation ability,technical influence and engagement.CNN-LSTM hybrid neural network was used to evaluate and predict the human resource value of software developers.The experimental results show that the accuracy of the model to evaluate the value of software developers can reach 98.59%,and the prediction results of human resource value of software developers on the test set are in line with the estimated value of enterprise recruiters.