Browsing Behavior Analysis from Wi-Fi Logs Based on Community Detection: Case Study on Educational Institution

Mei Silviana Saputri, Ari Wibisono, Adila Krisnadhi, Adhi Yuniarto Laurentius Yohanes, Teuku Amrullah Faisal, Alvin Wardhana Utama, Muhammad Azmi Malik Ariefa, Andre Ramadhani, Aminur Muda

Abstract: As the growth of technology, the need of internet increases. Wireless internet technology or called Wi-Fi is dominate. Hence, monitoring the usage of Wi-Fi is crucial. Community detection is a technique which can be used to understand browsing behavior analysis. Using this approach, a website is represented as a node in a graph and the transition over websites is represented as an edge where the weight of an edge is calculated based on the frequency of website transition. In this study, we implement two community detection algorithms which support directed and weighted graph, i.e Girvan-Newman and Infomap to know the communities of websites accessed by Wi-Fi users. An evaluation metrics named modularity is employed to access the quality of both algorithms. The results of our experiment show that Infomap performs better to detect communities than Girvan-Newman. Based on browsing behavior analysis, it can be understood that the Wi-Fi network is mostly used for education, entertainment and government purposes. However, we also obtain insight regarding the misbehavior activities which need prevention.