Graph Convolution Neural Network-based Revelation of Students' Career Expectations

Abstract

Clarifying the career expectations of freshmen is an important means to improve their future employment satisfaction. In the past, colleges and universities usually used questionnaires or personal interviews to understand students' career expectations. However, due to the complexity of everyone's goals and ideas, it is difficult to accurately predict students' professional expectations, and it is time-consuming and labor-intensive. With the development of data mining technology, many scholars use student behavior data to predict students' career expectations, but a large amount of data is needed for model training. Recent studies suggests that we can use the Career Planning Book assignment submitted by students to mine the career expectations of students. However, the career planning education of students in most domestic universities has not been carried out for a long time, and it is difficult to obtain a large amount of text data. Traditional data mining techniques, such as SVM and KNN, are not suitable for such small sample data, while Graph Convolutional Neural Network (GCN)model has a high accuracy for small sample data. Therefore, this paper proposes a model based on GCN, which can mine students' career expectations through Student's Career Planning Books. The experimental results show that GCN model can mine students' career aspiration with high accuracy under a circumstance of less data.