An Experimental Study of Factor Analysis over Cellular Network Data

TL;DR

This study applies exploratory factor analysis to cellular network data from Istanbul to uncover spatial and temporal behavioral patterns of residents, revealing commuting, cultural, and transportation behaviors.

Contribution

It introduces an empirical analysis of cellular network data using factor analysis to identify behavioral patterns, providing insights for network planning and user behavior understanding.

Findings

Identified morning and evening commuting factors.

Revealed cultural activity patterns like nightlife and weekends.

Discovered differences in transportation routes during rush hours.

Abstract

Mobile Network Operators (MNOs) are evolving towards becoming data-driven, while delivering capacity to collect and analyze data. This can help in enhancing user experiences while empowering the operation workforce and building new business models. Mobile traffic demands of users can give insights to MNOs to plan, decide and act depending on network conditions. In this paper, we investigate the behaviour of Istanbul residents using the cellular network traffic activity over spatial and temporal dimensions via exploratory factor analysis (EFA) using a major MNO's cellular network traffic data in Turkey. Our results reveal various time and spatial patterns for Istanbul residents such as morning and evening commuting factors, business and residential factors as well as nightlife and weekend afternoon factors as the most prominent cultural behaviour. The analysis results also demonstrate interesting findings such as tunnels and transportation paths selected by Istanbul residents may differ during morning rush work hour compared to evening rush after-work hour.