Recently, the Spatiotemporal Big Data and Intelligence Team from the School of Computer Science and Engineering (School of Cybersecurity) at the University of Electronic Science and Technology of China had three of their papers accepted by the IEEE International Conference on Data Engineering (ICDE2026), a leading conference in the database domain. The research spans topics such as next point-of-interest recommendation, semantic-aware publish/subscribe systems, and multi-dimensional point cloud data management. One of the papers introduces an efficient, model-agnostic continual learning framework named GIRAM. This framework boosts the model's ability to adapt to shifting interests through four key components. Experimental results demonstrate its superiority over existing methods in terms of both prediction accuracy and efficiency. Another paper presents a semantic-aware publish/subscribe framework called DISPS. This framework integrates a deep semantic embedding model with a collaborative hierarchical indexing structure, substantially enhancing the precision and robustness of topic matching. A third paper outlines a multi-dimensional point cloud data publish/subscribe system, PC-PS. This system guarantees low latency and dynamic updates in data-query matching by employing mechanisms such as constructing a reverse bounding box hierarchical index.