Abstract
Specialized services and management must understand students' behavioral patterns in a timely and accurate manner. Based on these patterns, we can make targeted rules, especially for unexpected patterns. To perform this type of work, a questionnaire method is usually used to collect data and analyze students' behavioral states. However, the effectiveness of this method is greatly influenced by the timeliness and validity of the feedback data. To address this problem, we propose an unsupervised ensemble clustering framework to use student behavioral data to discover behavioral patterns. Because the behavioral data produced by students on campus are available in real time without intentional bias, clustering analysis can be relatively efficient and reliable. The proposed framework extracts behavior features from the two perspectives of statistics and entropy and then combines density-based spatial clustering of applications with noise (DBSCAN) and k-means algorithms to discover behavioral patterns. To evaluate the performance of the proposed framework, we carry out experiments on six types of behavioral data produced by undergraduates in a university in Beijing and analyze the relationships between different behavioral patterns and students' grade point averages (GPAs). The results show that the framework can not only detect anomalous behavioral patterns but also find mainstream patterns. The findings from this research can assist student-related departments in providing better services and management, such as psychological consulting and academic guidance.
Highlights
The clustering results of density-based spatial clustering of applications with noise (DBSCAN) with the given values of Eps and MinPts are shown in Figs. 6 and 7, where −1 is the label of the noise cluster, the normal clusters are labeled with numbers starting from 0, and the number of students in each cluster is above its bar
To explain the role of DBSCAN in the proposed method, we adopt the principal component analysis (PCA) method to reduce the dimensionality of the behavior feature space to two and use a scatter chart to visualize the clustering results
Note that the cumulative variance explained by the two selected components in Fig. 12(a) is 68.6%, so they can basically express the distribution of students in the original space
Summary
An important task in the education field is discovering student behavioral patterns and taking the corresponding actions to optimize the educational process—for example, finding various behavioral factors that have strong correlations with academic performance [1]–[6], analyzing student learning behaviors to allow teachers to adjust teaching schedules for better outcomes and to give early warnings to students who may fail exams [7]–[10], modeling the mobility flow of students on campus to support the reasonable allocation of resources by administrators, detecting students’ anomalous behaviors so that managers can take timely preventive measures, and determining social networks from behavioral. Rich expert knowledge is needed to design a questionnaire that can capture enough information to comprehensively analyze students’ behavioral patterns These limitations make this method of data collection inefficient and costly. The proposed framework has four stages: data collection, feature extraction, clustering analysis, and visualization and evaluation. Six types of behavioral data produced on campus were collected from different information management systems using the extract-transform-load tool. In the final clustering results, the students that constitute noise and the students in small clusters discovered by DBSCAN can be considered anomalous, and the large clusters represent mainstream behavioral patterns. 1) Six types of behavioral data in time series format were collected, and the features for each type of behavior are extracted from the perspectives of central tendency, dispersion and entropy, which provides a more reliable basis for the analysis of behavioral patterns.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
