Impact of the spatial context on human communication activity

TL;DR

This paper explores how geographical context influences human communication activities by analyzing mobile data and geographic information, revealing patterns that can inform societal decision-making.

Contribution

It introduces a clustering approach combining spectral and k-means algorithms to identify activity-based geographic areas from mobile data and geographic info.

Findings

Clusters align with activity types like working and shopping

Silhouette coefficients validate cluster quality

Communication activity patterns vary across different geographic areas

Abstract

Technology development produces terabytes of data generated by hu- man activity in space and time. This enormous amount of data often called big data becomes crucial for delivering new insights to decision makers. It contains behavioral information on different types of human activity influenced by many external factors such as geographic infor- mation and weather forecast. Early recognition and prediction of those human behaviors are of great importance in many societal applications like health-care, risk management and urban planning, etc. In this pa- per, we investigate relevant geographical areas based on their categories of human activities (i.e., working and shopping) which identified from ge- ographic information (i.e., Openstreetmap). We use spectral clustering followed by k-means clustering algorithm based on TF/IDF cosine simi- larity metric. We evaluate the quality of those observed clusters with the use of silhouette coefficients which are estimated based on the similari- ties of the mobile communication activity temporal patterns. The area clusters are further used to explain typical or exceptional communication activities. We demonstrate the study using a real dataset containing 1 million Call Detailed Records. This type of analysis and its application are important for analyzing the dependency of human behaviors from the external factors and hidden relationships and unknown correlations and other useful information that can support decision-making.