Machine learning optimization for vocational literacy education evaluation: A big data-powered decision support system

This paper is motivated by the urgent requirement for creative solutions to address the challenges faced by vocational colleges in China’s rapidly advancing higher education system. It aims to use big data and data mining to improve vocational education and develop students’ professional characteristics. This study developed a comprehensive evaluation system for vocational education by using a decision support system (DSS) and data mining approaches based on big data analysis. The development is carried out in several stages. First, a complete DSS-based evaluation index system is developed by using large-scale data analysis. For this purpose, eight indicators were chosen to test students’ vocational literacy and create DSS and parameter matrices. Secondly, it uses the technique for order of preference by similarity to the ideal solution (TOPSIS) approach to analyze the results for each indicator, offering a solid foundation for decision-making. Thirdly, it uses regression analysis through the logistic regression model to investigate the particular features that impact students’ vocational literacy in vocational schools. Fourthly, the classification analysis is carried out to predict and analyze the vocational literacy level of vocational college students by using support vector machine (SVM), logistic regression, and AdaBoost. According to the assessment findings, 57% of students are judged competent or extremely competent, indicating that the majority have the essential vocational literacy for work. However, a comparison of students’ self-perceptions with enterprise ratings needs to be more consistent. Students tend to rank their vocational literacy better, with ratings around 4.0, whilst enterprise assessments linger around 3.6 points.