Dynamic Analysis and Forecasting of Employment Market Trends of College Graduates under Big Data Environment

Abstract

Abstract Predictive analysis of the job market before the employment of college graduates provides a direction to improve the quality of college students’ employment and better articulates the enrollment, training, and employment of colleges and universities. This paper utilizes the missing value filling algorithm and the ADASYN algorithm to preprocess the collected data related to the job market and graduates. Based on the HMIGW algorithm, a data feature selection method is proposed, and the feature collection is obtained after the feature selection process of each data series. Then, the XGBoost algorithm is used to predict and analyze the employment market trend. It was found that the accuracy of the model in analyzing and predicting the dynamics of employment market trends among college graduates reached 97.105% on average. After predicting the employment trend of graduates of a certain major in college S, it is found that the proportion of graduates who enter the Internet industry to work between 2020 and 2030 is relatively high, and the overall employment change trend of the agriculture and forestry job market is large. This paper can provide college graduates with a reference for choosing a career, and the predicted employment market trend can provide support for the development of employment guidance in colleges and universities. In future research, we can consider expanding the scope of the study and optimizing the prediction model.