基于注意力机制的兴趣点预测算法的研究

With the continuous development of information technology, more and more location-based services have emerged on the Internet. For example, intelligent transportation, logistics distribution, and unmanned driving. As an important subtask in location services, POI prediction, accurately predicting a user\'s next location by analyzing their movement patterns is crucial for improving location services and enhancing user experience. However, existing prediction algorithms based on recurrent neural networks cannot capture long-term dependencies when processing long sequence data. Therefore, this paper proposes a point-of-interest prediction model LSPAMbased on attention mechanism to capture users\' short- and long-term preferences with attention while capturing global contextual information without losing local information. In addition, prompt vectors and long-short term fusion strategies are introduced to further improve the performance of the model. Experiments show that the proposed method outperforms several current state-of-the-art algorithms.