Evaluating Weather Influence on User Participation in a Crowd-sensing Application.

The quality of life for people living in urban centers has become more difficult due to the rapid increase in the urban population. Crowdsensing services can be a helpful tool for improving people’s quality of life by giving public agents valuable data. This work proposes an analysis of traffic issues (events) reported by users of a crowd-sensing app in Vitória-ES, Brazil. The dataset was analyzed concerning the duration of the reported events, monthly variation of events, and the correlation between weather changes and events reported by users. Our findings indicate that the events exhibit an unclear seasonality pattern and may be influenced by climatic factors. We also found positive correlations between weather and events, especially when analyzing months individually. These results can provide new ways of investigation, e.g., new services for forecasting traffic problems.