Prediction and Influencing Factors of College Students' Career Planning Based on Big Data Mining

Abstract

College students' career planning refers to the process of establishing their own career development goals based on the analysis of their own subjective and objective factors, searching for a career, and taking necessary actions to achieve their career goals. Career planning is an important aspect of human resource management, and it is a new trend in recent research to use the theory and method of psychology to study it. This project understands the current situation of contemporary college students' career planning through investigation and finds the common problems in career planning among college graduates. Systematic analysis of these problems can be used as the research premise to study college students' career planning, help college students to make correct self-assessment and career orientation, improve their self-awareness and evaluation ability, and form a good career self. In the research, the author found that, in the guidance of college students' career development direction, it is necessary to predict the future career development direction of college students. Effective prediction methods can provide objective and simple data support and theoretical basis for educators to help them implement education guidance smoothly. Therefore, this paper has carried out a series of research work. The author firstly analyzes the problems existing in the process of college career education, then analyzes the relevant theories and common methods of educational data mining and the dimension analysis of the influencing factors of college students' career development direction, so as to determine the research dimension of this paper, and then carries out the relevant research such as the modeling of prediction model, the implementation of algorithm, and the case test of prediction method. According to the above research, the author draws the following conclusions: the grey prediction algorithm has better performance in small- and medium-sized data prediction. Based on the comprehensive quality evaluation data, the prediction conclusion is more objective and accurate. However, there are still some shortcomings in the algorithm design in this study. In the future research, we can try to adjust the relevant parameters through more data and more experimental verification, so as to further optimize the algorithm.