Points of interest idintification: A case study in Beijing metropolitian area

Dies ist die aktuellste Version dieses Eintrags.

Kurzfassung

Recent advances in navigation have generated an extensive amount of georeferenced trajectories which contain information about citizen's movement and their behavior. The objective of this article is to explore georeferenced satellite-based trajectories to identify points of interest, like university, restaurant and gym. We proposed a three-step approach, the first step is cleaning the georeferenced satellite-based data, the second step is to identify stay points by checking the distance and time between the points of each trajectory, and finally we cluster these stay points to extract points of interest. We evaluate the proposed method with a real-world dataset recorded over three years by 62 users in Beijing. We worked with trajectories where the users were walking, riding a bike and driving a car. Our approach was able to extract 138 points of interest, in which 40% were extracted from walking trajectories 33% from biking and 26% from driving. Our results have demonstrated that taking into consideration, the transportation mode for points of interest identification is an important aspect specially for the parameters estimation.

Verfügbare Versionen dieses Eintrags

• Points of interest idintification: A case study in Beijing metropolitian area. (deposited 23 Aug 2022 13:48) [Gegenwärtig angezeigt]