# 孤僻学生发现方法研究Research on the Method of Finding out Unsociable Students

• 全文下载: PDF(3081KB)    PP.637-648   DOI: 10.12677/CSA.2018.85072
• 下载量: 516  浏览量: 1,715

The goal of this article is to find students who are unsociable by analyzing students’ dining data. The concept of co-occurrence is introduced to calculate the number of students' co-occurrence and the time interval of co-occurrence, quantify student-to-student relationships, calculate student-relationship matrix, and use relational matrix values as distances to perform agglomerative hierarchical clustering. The accuracy rate of the predicted unsociable students was 66.67%, and the recall rate was 75%.

1. 引言

2. 学生就餐数据预处理

3. 学生就餐数据分析

Table 1. Sample of students’ dining data

Table 2. Sample of students’ conversional dining data

3.1. 学生间共现次数与学生是否孤僻之间的关系

Figure 1. 2013 students’ times of co-occurrence for the first semester

Figure 2. 2013 students’ times of co-occurrence for the second semester

Figure 3. 2013 students’ times of co-occurrence for the school year

3.2. 学生间共现时间间隔与学生是否孤僻之间的关系

Figure 4. 2013 students’ exponent of co-occurrences for the first semester

Figure 5. 2013 students’ exponent of co-occurrences for the second semester

Figure 6. 2013 students’ exponent of co-occurrences for the school year

Figure 7. 2013 students’ mean time interval for the first semester

Figure 8. 2013 students’ mean time interval for the second semester

Figure 9. 2013 students’ mean time interval for the school year

4. 学生关系模型的建立与应用

4.1. 学生关系模型的建立

Figure 10. 2013 students’ minimum time interval for the first semester

Figure 11. 2013 students’ minimum time interval for the second semester

Figure 12. 2013 students’ minimum time interval for the school year

$R\left(i,j\right)=tc\left(i,j\right)/count\left(i\right)/meant\left(i,j\right)/\mathrm{min}t\left(i,j\right)$ (1)

4.2. 层次聚类在学生关系模型上的应用

1) 将每个对象看作一个初始簇；

2) 根据距离度量标准找到最紧密的两个簇，合并，生成新的簇的集合，删除被合并之前的两个簇。

3) 重新计算新的簇和旧的簇之间的距离；

4) 重复2)、3)过程，直到簇的数目到k。

5. 实验结果

Table 3. The results summary table for the shortest distance

Table 4. The results summary table for the average distance

Table 5. The results summary table for the shortest distance + average distance

$\text{F}=2\text{PR}/\left(\text{P}+\text{R}\right)$ (2)

6. 结束语

 [1] 童睿. 基于一卡通消费数据的学生成绩预测和朋友关系网络检测研究[D]: [硕士学位论文]. 武汉: 华中师范大学, 2016. [2] 鲁鸣鸣, 张丹, 王建新. 基于校园一卡通数据好友发现及应用[J]. 大数据, 2017, 3(2): 78-91. [3] Eagle, N., Pentland, A. and Lazer, D. (2009) Inferring Friendship Network Structure by Using Mobile Phone Data. Proceedings of the National Academy of Sci-ences of the United States of America, 106, 15274-15278. https://doi.org/10.1073/pnas.0900282106 [4] Crandall, D.J., Backstrom, L., Cosley, D., et al. (2010) Inferring Social Ties from Geographic Coincidences. Proceedings of the National Academy of Sciences of the United States of America, 107, 22436-22441. https://doi.org/10.1073/pnas.1006155107 [5] Xu, B., Chin, A., Wang, H., et al. (2011) Social Linking and Physical Proximity in a Mobile Location-Based Service. International Workshop on Mobile Location-Based Service, ACM, 99-108. [6] Eagle, N., Pent-land, A. and Lazer, D. (2009) Inferring Friendship Network Structure by Using Mobile Phone Data. Proceedings of the National Academy of Sciences of the United States of America, 106, 15274-15278. https://doi.org/10.1073/pnas.0900282106 [7] Cutting, D.R., Karger, D.R., Pedersen, J.O., et al. (1992) Scatter/Gather: A Clus-ter-Based Approach to Browsing Large Document Collections. 318-329. [8] Zhou, J., Chen, C.L.P., Chen, L., et al. (2014) A Col-laborative Fuzzy Clustering Algorithm in Distributed Network Environments. IEEE Transactions on Fuzzy Systems, 22, 1443-1456. https://doi.org/10.1109/TFUZZ.2013.2294205 [9] Voorhees, E.M. (1986) Implementing Agglomerative Hierarchic Clustering Algorithms for Use in Document Retrieval. Information Processing & Management, 22, 465-476.