Location-based social networks (LBSN) are web platforms that reflect structures of social networks in the real world. In recent years, the study of LBSNs has attracted the attention of the scientific community, specially, because LBSNs consider information from user interactions and geographic location information for a period time, to develop applications such as location recommender systems, local travel planning systems and others. A specific problem in social network analysis is the exploration of the user dynamic behavior. In the context of LBSNs, the different proposals that explore user behaviors address only one problem, i.e., identification of dynamic patterns in user behavior, leaving open the exploration of two others problems: prediction of future structural changes and detection of unusual transitions in behaviors. Motivated by this gap, this project aims to investigate innovative techniques to effectively explore the various issues surrounding the user behaviors, considering the location history in a dynamic time domain. The scope of this proposal covers all stages of modeling a LBSN as well as the creation of a model of behavior based on the location information of users. The dynamic behavioral model to be developed in this project will build on the basis of existing behavioral models for traditional social networks. The results of this project will be validated in datasets available in the community of social networks analysis and data mining.
News published in Agência FAPESP Newsletter about the scholarship: