Exploratory Analysis Of The Data Of College Student Education Satisfaction Survey

Satisfaction is an indicator used to describe the subjective evaluation of investigators.It is generally used to describe the sense of pleasure after consumer demand is met,and the relationship between the expectations before the product or service and the actual feelings after use.In recent years,colleges and universities have increasingly investigated the satisfaction of students.The purpose is to understand the opinions and attitudes of college students on the construction and teaching of colleges and universities.This will help strengthen the communication between schools and students and enhance mutual understanding.At the same time,it is beneficial to college construction.Therefore,it has a strong practical significance to analyze the university student's school satisfaction survey data.In this paper,the results of a survey of satisfaction of all undergraduate students conducted by Yunnan University in 2018 are selected.First,the data are described and analyzed by using descriptive methods such as lists,related analy'sis,and statistical charts.Set up the structural equation model,analyze the factors that affect the satisfaction,and get the specific relationship between the observation variables(questionnaire results)and the latent variables(teacher job satisfaction,administrative service job satisfaction,professional curriculum satisfaction,and campus facility satisfaction).Clarify the significant effect of the observed variables on satisfaction.After that,according to the influence of different factors on the satisfaction,the overall satisfaction score is derived and calculated,and then combined with the question and answer session in the questionnaire,the specific proposal content of the question and answer questions is analyzed in text.Finally,the comprehensive satisfaction score and specific suggestions,respectively,use random forest,support vector machine,k-Nearest Neighbors and other machine learning methods to try to make a specific classification of students'personality preferences,divided into perfect,peaceful,active,and negative types.There are four types,Further dig into the survey data hidden in the subject's disposition information.The results of comparative analysis show that the classification of random forests is the best.The research feature of this paper is to carry out exploratory data analysis of survey data in combination with machine learning method.Based on the traditional method of using structural equation model to analyze survey data,and combining with machine learning method to deeply excavate data information,making the investigation data use more fully.This study provides a methodological reference for the university to analyze and process the student satisfaction survey data.At the same time,the research results also help the school to strengthen the understanding and management of students,conduct better communication with students,and establish a more complete training system.