Campus Daily Behavior Patterns: A Research Based On Big Data And Personality Model

As an emerging technology, big data has strongly advanced the development of web science in the age of information. The penetration of big data in various sectors bring us huge benefits to our lives. However, it also brings new challenges at the same time. In terms of education, the society with information explosion requires the education organizations that can not only manage the rapid growth of students’ information, but also mine valuable information from the data to guide the students for better learning and social experiences. From the perspective of promoting the research of big data in the field of education, this dissertation first expounds the importance of understanding the campus daily behavior patterns. Based on the personality model in psychology, this dissertation then interprets and classifies the daily behavior patterns further. Finally, combines with an anonymous data base that covers the students’ information of UESTC, this dissertation does three research works, which are(1) the correlation between the students’ campus behavior patterns and academic performance,(2) the intrinsic correlation between different behavior features, and(3) the correlation between behavior patterns and regions.The research work and contribution of this dissertation mainly contains three parts that are organized as follows.1.In terms of the situation that education organizations cannot judge the alternative trends of students’ performance before the exam, this dissertation tries to solve it by taking the narrow traits of Conscientiousness in the Big Five Personality model.It analyzes and selects the study-related behavior features based on the narrow traits Achievement and Order respectively. According to the behavior features selection, it then grabs and quantifies the corresponding behavior records and academic performance records of huge amount of students from the data base. Cumulative times and actual entropy are applied to quantify the discrete behavior data,while the normalized rank is used to represent the academic performance. Then a correlation analysis is done between the behavior features and the normalized rank. The result shows that obvious correlation exists between the Achievementrelated behavior features and normalized rank. And the correlation between the Order-related behavior features and the normalized rank also cannot be neglected.2.The narrow trait Order reflects the tendency of apply structure to work and life.More than one campus daily behavior features corresponds to it. To solve the problem that to which degree the intrinsic correlation exists between the different behavior features, this dissertation adopts the conditional probability and Kendall rank correlation coefficient to compute the static and dynamic intrinsic correlation between the Order-related behavior features. The result indicates that among all Order-related behavior features, meals in cafeterias has the strongest intrinsic impact. In addition, the intrinsic correlation of all the behavior features are strengthen as time goes.3.Consider that the patterns of Order-related behavior features of students may have huge differences, this dissertation proposes a hypothesis that the differences may have correlation to the regions, specifically, to the hard level of College Entrance Examination of different regions. It then looks for the information of College Entrance Examination scores over a 2-year period according to the students’ native places, and converts the scores to hard level rank of enrolling in UESTC.To avoid the potential impact to the hard level caused by environment, it selects the geographically adjacent region couples with huge hard level gap. Correlation analysis is done between the selected region couples and Order-related behavior features. The result illustrates that the hard level of College Entrance Examination has weak negative correlation, no correlation, and positive correlation with the behavior features of library check-in, shower in dormitories, and meals in cafeterias respectively.