Abstract
Human capital is one of the critical resources for high-tech industries such as semiconductor manufacturing to maintain their competitive advantages, yet it is seldom addressed in literature. Owing to the changing nature of knowledge workers in high-tech industries, jobs cannot be easily delineated. Thus, conventional personnel selection approaches based on static job characteristics no longer suffice. Focusing on the needs in real settings, this study aims to develop a manufacturing intelligence framework that integrates the rough set theory, support vector machine, and decision tree to extract useful patterns and intelligence from huge human resource data and production data to enhance the decision quality of human resource management that include identifying high-potential talents who fit the company culture and allocating the job with functional nature that matches the characteristics of the talent. To assess the validity of this approach, empirical studies were conducted on the basis of real data collected from semiconductor companies for comparison. The results have shown the practical viability of this approach. © Springer Science+Business Media, LLC 2010.