Article: Prediction of talent demand and job matching based on knowledge graph and attention mechanisms Journal: International Journal of Information and Communication Technology (IJICT) 2024 Vol.25 No.9 pp.76 - 87 Abstract: The real-time job data from recruitment platforms, reflecting the demands of enterprises for job seekers, can provide data support for the development of university training policies. This study proposes a knowledge representation model based on the self-attention mechanism for talent demand and job matching prediction method, where neural networks aggregate neighbourhood information to generate better node representations. In doing so, nodes can learn their local neighbourhood information and the whole graph structure. At the same time, using the self-attention mechanism, further extracts node features containing rich neighbourhood information. Deep interactions between nodes are used to calculate the central entity around the attention coefficients of neighbours to improve the accuracy of prediction. The experimental results show that the model prediction is highly fault-tolerant, short time-consuming, and can be widely used in the practical application of matching talent demand and university talent cultivation. Inderscience Publishers - linking academia, business and industry through research

Title: Prediction of talent demand and job matching based on knowledge graph and attention mechanisms

Authors: Xiaoli Mei

Addresses: School of Management, Jiangxi University of Technology, Jiangxi 330098, China

Abstract: The real-time job data from recruitment platforms, reflecting the demands of enterprises for job seekers, can provide data support for the development of university training policies. This study proposes a knowledge representation model based on the self-attention mechanism for talent demand and job matching prediction method, where neural networks aggregate neighbourhood information to generate better node representations. In doing so, nodes can learn their local neighbourhood information and the whole graph structure. At the same time, using the self-attention mechanism, further extracts node features containing rich neighbourhood information. Deep interactions between nodes are used to calculate the central entity around the attention coefficients of neighbours to improve the accuracy of prediction. The experimental results show that the model prediction is highly fault-tolerant, short time-consuming, and can be widely used in the practical application of matching talent demand and university talent cultivation.

Keywords: knowledge graph; attention mechanism; talent demand; job matching.

DOI: 10.1504/IJICT.2024.143327

International Journal of Information and Communication Technology, 2024 Vol.25 No.9, pp.76 - 87

Received: 27 Sep 2024
Accepted: 11 Oct 2024
Published online: 13 Dec 2024 *

Title: Prediction of talent demand and job matching based on knowledge graph and attention mechanisms

Keep up-to-date