Unveiling urban and human mobility dynamics through semantic trajectory summarization

The analysis of semantic trajectories has gained significant attention in urban mobility research due to its potential to provide comprehensive insights into movement patterns and associated semantic data. However, integrating multiple semantic aspects with spatio-temporal information often leads to redundancy and computational challenges. To address this issue, we proposed MAT-SUM, a novel method for trajectory summarization while preserving semantic quality. This method identifies urban regions based on associated semantic contexts and discretizes trajectories, yielding a high level of summarization without compromising semantic quality. This paper delves into the research questions we aim to address, focusing on the effectiveness of capturing human mobility and urban dynamics following the summarization of semantic trajectories using MAT-SUM. Our investigation aims to determine if comparable results can be achieved by analyzing summarized semantic trajectories compared to the original ones, such as classifying users exhibiting routine and non-routine behaviours. Furthermore, we examine the usefulness of the identified regions in comprehending urban dynamics. While some research questions remain open, this paper outlines ongoing investigations and potential strategies to enhance outcomes in mobility analysis.