Social network analysis based on canteen transaction records (Q1782106)

This paper studies social network analysis for college students' social interaction building on canteen transaction records. The premise is that two students who frequently purchase meals congruently in the same sequence are more likely to be socially connected. The friendship relation in canteen is determined by using a mechanism to determine the threshold frequency of such queue contiguous event. The discovered relationship is a strong correlation with respect to students' behavior compared to some static community. The students' position in the network can be found by calculating the relevant attributes. As a numerical study, a canteen transaction data set is studied, which has around 4.5 million canteen records. The data contains around 16,000 undergraduate students' canteen transaction records during one semester. The method proposed in this paper has been well illustrated in the numerical study.