Weather Effects on Mobile Social Interactions: A Case Study of Mobile Phone Users in Lisbon, Portugal

Abstract

The effect of weather on social interactions has been explored through the analysis of a large mobile phone use dataset. Time spent on phone calls, numbers of connected social ties, and tie strength were used as proxies for social interactions; while weather conditions were characterized in terms of temperature, relative humidity, air pressure, and wind speed. Our results are based on the analysis of a full calendar year of data for 22,696 mobile phone users (53.2 million call logs) in Lisbon, Portugal. The results suggest that different weather parameters have correlations to the level and character of social interactions. We found that although weather did not show much influence upon people's average call duration, the likelihood of longer calls was found to increase during periods of colder weather. During periods of weather that were generally considered to be uncomfortable (i.e., very cold/warm, very low/high air pressure, and windy), people were found to be more likely to communicate with fewer social ties. Despite this tendency, we found that people are more likely to maintain their connections with those they have strong ties with much more than those of weak ties. This study sheds new light on the influence of weather conditions on social relationships and how mobile phone data can be used to investigate the influence of environmental factors on social dynamics.

Predicting a Community's Flu Dynamics with Mobile Phone Data

“…More generally, behavior inferred by mobile sensing has recently been shown to bear a relationship with changes in weather [19], political opinions [15,4], and personality traits [3]. These promising advances in modeling human behavior reveal that mobile sensing may be a tool which could eventually address challenges currently facing epidemiology research, in term of both data quality as well as scenario complexity.…”

Deriving human activity from geo-located data by ontological and statistical reasoning

“…Therefore, associating environmental and social factors to mobile phone data is extremely useful for analyzing the dependency of human behaviors from the external factors. These studies focus on the inference of a certain level of human activities under certain conditions like human behavior changes in uncomfortable weather conditions [30,46,46,65], human mobility of different communities [10,45], human mobility during an event [2,8,42], or communication activity patterns in different land-use types [24,44,45,60]. In this work, geographical information in particular points of interest (POIs) are collected using pYsearch (Python APIs for Y!…”

“…For every record in a CDR dataset, it is possible to determine to a certain level of approximation the context associated to such a record. Some research studies [24,30,45,46,46,65] have been performed to understand the correlations of human behaviors to environmental factors, using some additional contextual information about weather, social events and geographical information systems. This focuses on the inference of a certain level of human activities in a certain condition.…”

Socio-inspired ICT

“…The authors also investigate patterns of calling activity at the individual level and model the individual calling patterns (time between phone calls) as heavy tailed. The most recent work considering a very large scale mobility dataset obtained upon phone call initiation is by Phithakkitnukoon et al [121], where they study the correlation between weather patterns and mobile phone usage.…”