A spatial influence query aims to find the influence zone or influence objects (e.g., customers) of a spatial facility (e.g., a store outlet) with respect to certain spatial criteria (e.g., reverse nearest neighbor). Owing to its broad applications in decision support, market analysis, social media advertising and recommender systems, the problem of spatial influence queries has been studied extensively in the literature. However, most of existing work focuses on spatial proximity only. With the rapid development of location- based services (LBS), a large number of spatial facilities/objects have been enriched with text information (e.g., categorical descriptions or user-generated reviews). So far, little attention has been paid to spatial influence queries for such spatial-textual facilities/objects. In this project, we propose a novel query paradigm -- keyword-based spatial influence query -- to retrieve the influence zone or influence object set of a spatial-textual facility on the basis of both spatial proximity and textual similarity. Specifically, we plan to study three representative keyword-based spatial influence queries: 1) keyword-based influence region search; 2) keyword-based influence set search; 3) maximizing keyword-based influence region/set search. We show the technical challenges of processing these queries. We plan to develop new index-based search methods and approximate algorithms for fast query evaluation. Cost models and query optimization techniques will also be investigated. In addition, we plan to build a map-based demonstration system to showcase the algorithms and techniques developed in this project.
|Effective start/end date||1/01/18 → 31/12/20|
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.