Construction and Practical Research on the Dynamic Analysis Model of College Students' Employment Data in Higher Education Institutions under Big Data Environment

Abstract

In the context of big data, college students face numerous challenges and opportunities in terms of employment. The processing of employment dynamic data suffers from issues such as poor employment trend prediction and weak precision in employment services. To address these issues, this paper employs big data to quantitatively process employment data. Using the K-means clustering algorithm, the paper determines the number of clusters in the employment data and completes the data clustering process. It primarily calculates the degree centrality parameters of network nodes to identify key points in the employment data required by users, thereby achieving employment data visualization. Based on time series analysis methods, the paper constructs a dynamic analysis model for college student employment data to analyze and predict employment trends among college students. Employment trend prediction practices are conducted using employment data from a certain higher education institution from 2017 to 2024 to predict the employment trends of college students in 2025. The employment index for the coming year in 2025 continues to show cyclical fluctuations, peaking in the fourth week (spring recruitment period), then rapidly declining, and recovering and rising to its highest point after the 30th week (autumn recruitment period). The results of the employment trend prediction generally align with actual social conditions.