摘要： 针对采用欧式空间坐标或线性参考坐标表达轨迹的方法对与应用需求相关的背景地理语义知识考虑不够，使得不能有效挖掘轨迹中隐含信息的问题，提出一种基于语义的轨迹建模和知识发现方法，该方法通过预先定义与应用需求相关的重要地点作为关键点，对与该地点具有同种空间关系的轨迹点进行聚类，以聚类后的关键点序列表达轨迹，运用正则表达式实现轨迹中隐含的关联规则和频繁模式的发现，通过实例表明该方法的可行有效。

关键词: 轨迹, 语义, 知识发现

Abstract: Due to the traditional trajectory model that composed of the Euclidean or linear reference coordinates sequence can not present the implied semantic information of the background geographic information, it impedes the knowledge discovery on the trajectories. The paper proposes a semantic based trajectory knowledge discovering method. It clusters the original GPS x, y coordinates that have the same spatial relation with the predefined key point of the application into one key area at first. The trajectory is depicted using the clustered key area sequence. The regular expression is used to analyze the trajectory. Experiment shows it contributes to simplify the knowledge discovery process.

Key words: trajectory, semantic, knowledge discovery

中图分类号: 

• TP311