Human Resource Matching Support System Based on Deep Learning

Abstract

Aiming at the problem of reasonable recommendation and accurate matching of human resources, a hybrid human resources matching recommendation algorithm based on GBT-CNN is proposed in this article. The advantages of traditional GBT and CNN algorithms are combined and can give full play to the high-level feature abstraction ability of convolution processing. The gradient lifting tree is used to transform the features, complete the feature screening and coding, and then input the hybrid convolution neural network to obtain the high-dimensional feature abstraction by using the hybrid convolution operation, to improve the quality of human resources recommendation. In this article, GBT and CNN algorithms are first described, and then the basic framework and specific implementation of GBT-CNN algorithm are introduced. Finally, the effectiveness of the algorithm is verified by simulation experiments. The results show that the deeper correlation between job seeker information and job information can be effectively captured by the algorithm. Besides, reasonable recommendation and accurate matching of human resources can be realized.