Analysis And Research Of Personality Tendency Based On College Students Network Behavior Data

At present,most students have insufficient cognition of the pathological state of personality defects.They are not found in time and have low willingness to treat.Defective personality may even lead to marginal psychology and anti-social psychology.Therefore,it is important to know whether the student’s personality is healthy in time.The amount of data generated by college students online has increased dramatically.These data provide us with unprecedented opportunities to track,analyze and predict student personality.In particular,recent advances in data mining technology have not only provided us with strong support from data,but also provided more scientific analysis methods.At present,many researchers have conducted personality-related research.These researches can be roughly divided into two methods,namely,the traditional questionnaire method and the research method combined with big data.However,the traditional questionnaire method is time-consuming and laborious,and the reporting deviation is large.The reason for this result is that the questionnaire method is poor in timeliness and difficult to track continuously for a long time.There are few behavioral performance studies,and there are few specialized studies on student groups.The reason may be that researchers have not paid enough attention to other online behaviors or cannot obtain other online behavior information.In response to these problems,this research proposes a method for predicting students’ personality orientation based on the Internet behavior data generated by students surfing the Internet at school,combined with the traditional scale method.This method can analyze whether the students’ personality orientation is healthy in real time.This research expands the mining object data to a wider range of behavioral information including social network behaviors,introduces correlation analysis and regression analysis techniques in data analysis methods into the research of traditional scale methods,and uses data from data mining Classification and data integration mine the behavioral data generated by students surfing the Internet at school.These data not only contain social network information from previous studies,but also other behavioral information of students.Then combine the analyzed scale method with the big data research method,and choose different algorithms to build the student personality tendency analysis model to analyze and predict the student’s personality tendency,so as to assist the school and parents to help students shape a healthy personality in time.This study elaborated and analyzed the collected data of the student assessment scale,and calculated the correlation between the measurements in different scales through the Pearson product-moment correlation coefficient,and found outliers,anxiety,and hostility.It has a great correlation with non-extraverted personality,and uses this as a theoretical basis to integrate the results of different scales.We divide the student sample data into static attribute data of basic attributes and dynamic attribute data of students’ surfing behavior,and carry out feature construction respectively.Then use mutual information method and logistic regression analysis method to measure the correlation between static data and personality tendency,and use recursive feature elimination and evolutionary algorithms to optimize dynamic features before input.In the study,a variety of different classification and integration models of core algorithms were constructed,and horizontal comparison experiments between different models were designed,and vertical comparison experiments under different parameters of the same model were designed.The experimental results show that there is indeed a strong correlation between student network behavior data and their personality tendencies.This conclusion is not only based on theory,but also supported by conclusions in data analysis.Therefore,the method proposed in this study to use students’ network behaviors to predict their personality tendencies is feasible.The model can be constructed to realize the judgment of students’ personality tendencies,thereby assisting schools and parents to help students shape healthy personality in time.