利用使用者軌跡偵測停留點

隨著科技的進步，配置 GPS 感測器的裝置已逐漸深入人們生活的習慣。藉由分析 GPS 路徑中所包含的位置資訊及時間戳記，我們可以從中找出使用者的地理和空間資訊並將其使用在適地性的服務中。在本篇研究中，我們將目標放在如何利用分析使用者的 GPS 軌跡來偵測停留點；之後利用偵測出的停留點觀察使用者的旅遊習慣和喜好。然而目前的相關研究所偵測的停留點皆較為粗略，造成停留點準確度下降進而影響後續的分析結果。因此本篇研究將著重於提升停留點的精確度和減少錯誤偵測的發生。為了達成以上的目標，首先我們先將 GPS 軌跡中的每個 GPS 點利用時間和距離的資訊計算相對速度，並以此為基準將 GPS 軌跡分割成數個部分後，再依照停留時間和平均速度來找出停留位置。接著藉由從所有軌跡中找出的停留位置的密集度得到相對應停留點的區域和核心點。之後我們利用核心點與網路上的興趣點列表比對，進而找出停留點真正的地點名稱和相關資訊。最後利用偵測到的停留點觀察使用者的旅遊類型偏好，並以此推薦適合的停留點給使用者。本篇的最後針對使用參數和相關研究成果比對進行各項實驗，而實驗結果顯示我們的研究成果有顯著的成效和優異的結果。

關鍵字

GPS軌跡分析；歷史位置資訊；時空資料探勘；適地性服務

並列摘要

The mobile devices with GPS sensors that have become widely available in recent years have changed the way people record their lives. By tracking users GPS trajectories which consist of geographical positions and timestamps, we can extract significant useful information to support location-based services. In this paper, we aim to discover stay locations that a user has visited while traveling by analyzing the GPS trajectories. As a result, a trip recommender system can effectively obtain his/her behavior (ie, traveling order) and preferences inferred from the characteristics of the stay locations. However, the existing work only adopts a rough region covering a set of GPS points on a trajectory for representing a stay location, incurring low precision analysis results. We focus on discovering precise stay locations using the following steps. Firstly, the stay points are extracted from the GPS trajectories and are used to divide the trajectories into segments based on the factors of the moving speed and the stay time. Secondly, the stay points are merged into one stay location based on their density. Subsequently, we identify the stay locations from a Point of the Interest (POI) list crawled from the Internet, and merge the stay locations which are identified as the same POI. Lastly, we apply the result to the recommendation system. The stay locations are extracted to obtain the user's preferences, and finally the locations which match the preferences are recommended to the user. We have conducted experiments for evaluation, and our method is shown to outperform the existing approaches.

並列關鍵字

無資料

參考文獻

[3] D. Ashbrook and T. Starner. Using gps to learn significant locations and predict movement across multiple users. Personal and Ubiquitous Computing, 7(5):275– 286, 2003.

[5] X. Cao, G. Cong, and C. S. Jensen. Mining significant semantic locations from gps data. Proceedings of the VLDB Endowment, 3(1-2):1009–1020, 2010.

[9] Y. Koren, R. Bell, and C. Volinsky. Matrix factorization techniques for recommender systems. Computer, (8):30–37, 2009.

[12] J. Liu, O.Wolfson, and H. Yin. Extracting semantic location from outdoor positioning systems. In MDM, page 73. Citeseer, 2006.

[15] R. Montoliu, J. Blom, and D. Gatica-Perez. Discovering places of interest in everyday life from smartphone data. Multimedia Tools and Applications, 62(1):179–207, 2013.

國際替代計量

利用使用者軌跡偵測停留點

未授權

主題瀏覽