Keresteci, E.Bulut, M.E.Akgun, M.B.Tavli, V.B.2025-03-222025-03-2220249798331518035https://doi.org/10.1109/ELECO64362.2024.10847099https://hdl.handle.net/20.500.11851/12137Spatio-temporal finger-printing is an active research topic, which is of critical importance in mobility analysis. Various neural network models working on LBSN data have been presented within the framework of TUL (Trajectory-user linking). However, to the best of our knowledge, there are no studies on neural networks for sequential spatio-temporal data like GPS in the literature. Therefore, here we present the STSeqGNN model, which can process sequential spatiotemporal data. Our model can process the graph structure of the map and the time dimension of the data. Our model also can use movement information and basic statistical values for better finger-printing performance. We present our test results with the evaluation metric accuracy at k. Our model achieves more than 99% accuracy across different datasets. © 2024 IEEE.trinfo:eu-repo/semantics/closedAccessConference Object2-s2.0-8521783711210.1109/ELECO64362.2024.10847099