支持语义信息挖掘的热点路径探测

Hot routes detection based on semantic information mining

针对现有热点路径探测算法存在缺乏对轨迹语义信息进行分析的问题,提出一种支持语义信息挖掘的热点路径探测算法:首先研究轨迹数据语义空间的建模方法,并据此构建低维语义子空间来计算轨迹数据语义相似度,描述轨迹所属移动对象的社会角色的相似性,最后结合基于轨迹流与轨迹密度的传统热点路径探测算法实现对不同社会角色对应的热点路径的发现。结果表明,该算法能够较好利用轨迹数据的时空和语义信息,有效识别出不同社会角色对应的热点路径的聚类特征,为个性化的位置服务研究提供参考。

Aiming at the problem that it is lack of the analysis on semantic information of trajectories in the existing algo- rithms of hot routes detection, the paper proposed a detection method supporting semantic information mining., firstly, the modeling method of semantic space of trajectory data was studied~ secondly, the low-dimensional semantic subspace was constructed to compute the semantic similarity which describes the comparability of the social roles of the moving objects; finally, combined with the traditional hot routes detection algorithm based on trajectory flow and density, the discovery of hot routes corresponding to different social roles was realized. Result showed that the proposed method could make use of the spatial-temporal and semantics information of the trajectory data, and effectively identify the clustering characteristics of the hot routes corresponding to different social roles, which would provide a reference for related study on personalized location-based services.