An Improved Apriori Algorithm for Association Rule Mining in Employability Analysis

The wide application of emerging advanced technologies causes significant changes in the development trend of the employment market. The lack of flexible and easy-to-implement analysis methods challenges general maritime education practitioners to understand the developing trends. This research proposed the improved Apriori algorithm to explore employment preference by identifying the association rule of the employability indicators and the employment status. The candidate generation methods are optimised based on the questionnaire design to generate fewer candidates. The minimum support value is automatically generated to reduce the reliance on analysis expertise and improve accuracy. To validate the algorithm, a questionnaire for the maritime graduate is used to collect employment data to test the efficiency and capability of the improved algorithm. The computation time for different data set sizes shows that the improvement could improve the algorithm's effectiveness. The algorithm also successfully identifies significant employment preference that certain employment types emphasise specific employability skills, such as responsibility and core professional skills. The results suggest that the improved A algorithm could reduce the computing burden and identify the employment preference from questionnaire data. This research provides easy-to-use and flexible analysis tools, which could reduce the computing expertise required for education practitioners.