Explore the Ubiquitous Learning on Campus: A Friendship-Based Knowledge Diffusion Approach

Interactions among people is one of the primary means to acquire knowledge on campus, smart campus systems help us to obtain a clear picture about how network structures influence the performance of knowledge transfer. In this paper, we develop a Wi-Fi hotspot-based mobile application that can collect the student's data by considering the mobile phone as information sensing device. The application collects the data including the partners, the times, and the duration of interactions among people without violating their privacy. Then construct friendship networks by processing the data. Moreover, a new knowledge diffusion model is presented to measure the knowledge spread quantitatively on different network structures, which is different from the epidemic model. Simulation results are provided to illustrate that the proposed is more to satisfy the actual situation. For this model, the knowledge diffusion performance of regular network is better than small-world networks and scale-free networks. Hence, the comparative experiments are made among three classes in two course in terms of the average scores and the standard deviations, the results verify the positive effect of our approach.