Title: Risk warning method for the whole process of production project based on multi-source data mining

Authors: Ya Zhou; Mingjie Zhang; Shaohuan Cheng; Xinli Luo; Jiefeng Wen; Jinlei Hu

Addresses: Qingyuan Power Supply Bureau of Guangdong Power Grid Co., Ltd., Qingyuan, 511500, China ' Qingyuan Power Supply Bureau of Guangdong Power Grid Co., Ltd., Qingyuan, 511500, China ' Qingyuan Power Supply Bureau of Guangdong Power Grid Co., Ltd., Qingyuan, 511500, China ' Qingyuan Power Supply Bureau of Guangdong Power Grid Co., Ltd., Qingyuan, 511500, China ' Qingyuan Power Supply Bureau of Guangdong Power Grid Co., Ltd., Qingyuan, 511500, China ' Qingyuan Power Supply Bureau of Guangdong Power Grid Co., Ltd., Qingyuan, 511500, China

Abstract: Aiming at the problem that the correlation coefficient between the final screening early warning indicators is small in the whole process of risk index screening of production projects, a risk early warning method for the whole process of production projects based on multi-source data mining is proposed. The multi-source data mining technology is used to extract the potential hidden characteristics of risk indicators in the whole process of production projects, and the hidden characteristic relationship between multi-risk indicators is obtained, and the risk management index system of the whole process of production projects is established. Risk warning of the whole process of production projects. The test results show that the proposed risk early warning method has a larger cross-correlation coefficient calculated in each index group, which verifies that the early warning method has high application reliability, can obtain more accurate data, and improve the early warning accuracy.

Keywords: multi-source data mining; production project; whole process risk; risk early warning; K-means cluster analysis algorithm; index correlation characteristics.

DOI: 10.1504/IJISD.2025.143053

International Journal of Innovation and Sustainable Development, 2025 Vol.19 No.1, pp.94 - 106

Received: 18 Aug 2022
Accepted: 31 Oct 2022
Published online: 02 Dec 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article