Extracting Mobile User Profile using Easy-to-obtain and Less Invasive Data

In recent years, we have been facing a significant increase in the interest of the research community and industry in geospatial data. Using this type of data to extract the behavior of mobile users benefits them when this knowledge is used on their behalf. For example, companies can offer personalized and location-based services based on the user's mobile profile. Thus, building a user profile regarding their behavior and mobile interests is paramount to improving the offered services and guiding future applications such as Intelligent Transportation Systems. However, mobile profiles require a large amount of geospatial data that is not always available or usually invades the user's privacy. To mitigate this problem, in this work, we investigate if it is possible to use the users' city and mobile apps to infer their mobility behaviors and consumption patterns. For this purpose, we created models based on four perspectives: device price, categories of visited venues, visited functional areas, and commuting patterns. The results revealed that our models can infer the users' mobile profiles with good precision and recall values, considering only less invasive information.