STF-RNN: Space Time Features-based Recurrent Neural Network for predicting people next location

Al-Molegi, Abdulrahman; Jabreel, Mohammed; Ghaleb, Baraq

Abstract

This paper proposes a novel model called Space Time Features-based Recurrent Neural Network (STF-RNN) for predicting people next movement based on mobility patterns obtained from GPS devices logs. Two main features are involved in model operations, namely, the space which is extracted from the collected GPS data and also the time which is extracted from the associated timestamps. The internal representation of space and time features is extracted automatically in the proposed model rather than relying on handcraft representation. This enables the model to discover the useful knowledge about people behaviour in more efficient way. Due to the ability of RNN structure to represent the sequences, it is utilized in the proposed model in order to keep track of user movement history. These tracks help the model to discover more meaningful dependencies and as consequence, enhancing the model performance. The results show that STF-RNN model provides good improvements in predicting people's next location compared with the state-of-the-art models when applied on a large real life dataset from Geo-life project.

Citation

Al-Molegi, A., Jabreel, M., & Ghaleb, B. (2017). STF-RNN: Space Time Features-based Recurrent Neural Network for predicting people next location. In 2016 IEEE Symposium Series on Computational Intelligence (SSCI). https://doi.org/10.1109/ssci.2016.7849919

• Conference Name: 2016 IEEE Symposium Series on Computational Intelligence (SSCI)
• Conference Location: Athens, Greece
• Start Date: Dec 6, 2016
• End Date: Dec 9, 2016
• Online Publication Date: Feb 13, 2017
• Publication Date: 2017
• Deposit Date: Jan 25, 2022
• Publisher: Institute of Electrical and Electronics Engineers
• Book Title: 2016 IEEE Symposium Series on Computational Intelligence (SSCI)
• DOI: https://doi.org/10.1109/ssci.2016.7849919
• Public URL: http://researchrepository.napier.ac.uk/Output/2837742