The proliferation of geo-social network, such as Foursquare and Facebook Places, enables users to generate location information and its corresponding descriptive tags. Using geo-social networks, users with similar interests can plan for social activities collaboratively. This paper proposes a novel type of query, called Tag-based top-k Collaborative Spatial (TkCoS) query, for users to make outdoor plans collaboratively. This type of queries aim to retrieve groups of geographic objects that can satisfy a group of users' requirements expressed in tags, while ensuring that the objects be within the minimum spatial distance from the users. To answer TkCoS queries efficiently, we introduce a hybrid index structure called Spatial-Tag R-tree (STR-tree), which is an extension of the R-tree. Based on STR-tree, we propose a query processing algorithm that utilizes both spatial and tag similarity constraints to prune search space and identify desired objects quickly. Moreover, a differential impact factor is adopted to fine-tune the returned results in order to maximize the users' overall satisfaction. Extensive experiments on synthetic and real datatsets validate the efficiency and the scalability of the proposed algorithm. © 2012 Springer-Verlag.
CITATION STYLE
Zhang, J., Meng, X., Zhou, X., & Liu, D. (2012). Co-spatial searcher: Efficient tag-based collaborative spatial search on geo-social network. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7238 LNCS, pp. 560–575). https://doi.org/10.1007/978-3-642-29038-1_40
Mendeley helps you to discover research relevant for your work.