Data Clustering Techniques to Identify User Groups and Resource Grouping in nanoHUB

With a massive increase in the number of online resources for education and research, it is important to study their usage by target audience comprised mainly of students, educators and researchers. This study explores the application of data clustering techniques on user access data of online science platforms in order to detect user groups and categorize resources with the aim of finding evidence that nanoHUB, the largest science gateway in the field of nanotechnology, aids educational advancement and research. Several algorithms are examined to find the best-suited algorithm for the data set in question. The study uses a two-stage methodology to find classroom like user groups with the help of clustering and further evaluates categorization of the set of resources used by such groups based on a limited set of available features. The techniques used in the methodology are Spatio-Temporal Density Based Scan to detect groups of similar users and Jaccard index to find resource categories by monitoring continued usage of nanoHUB by these groups of users. The resulting user groups and resource sets are evaluated to understand the utility of nanoHUB in a classroom-like group. From the resulting grouping, we can say that spatiotemporal clustering based on a limited number of features reveals group usage patterns of nanoHUB across the globe