Data-driven analysis for knowledge-action integration of university graduates

There are high lack of employment and the unemployment rate in Taiwan. The future employment situations of graduates become an objective of university. Employability is often used to evaluate knowledge, skills and attitudes of students (Pool and Sewell, 2007). In many universities, we continued import business practice, the industry teaching and other ways to improve students' employability.

This study aims to find key factors which affecting graduates from learning performance by using data-driven data mining and big data tools. There are three tasks in this study: (1) Define “knowledge-action integration” or not: Graduates will be evaluated by their “industrial” or “position.” (2) Integrate data of affecting factors: We will use different data sets from different sources. We need use data preprocessing, data clearing to integrate data sets. (3) Find the fittest model: We will select fittest model from several models, such as decision trees, regression, and so on.