Social-aware Spatial Keyword Top-K Group Query
Zhao Xiangguo,Zhang Zhen,Huang Hong,Bi Xin
DOI: https://doi.org/10.1007/s10619-020-07292-0
IF: 0.974
2020-01-01
Distributed and Parallel Databases
Abstract:With the increasing popularity of location-based social networking services, information in social networks has become an important basis for analyzing user preferences. However, the existing spatial keyword group query only focuses on the distance constraint between the user groups, and ignores the social relationship between the user and his friends, which may affect the query results. Therefore, in order to meet the diverse query needs of user groups and improve user satisfaction based on information in social networks, this paper proposes a social-aware spatial keyword top-k group query problem. This problem aims to retrieve a set of k groups of POI objects that satisfy the preferences of multiple users, taking into account spatial proximity, social relevance, and keyword constraints. To solve this problem, we first design a rank function to measure the correlation between the query set and the candidate set. Next, in order to improve the query efficiency, we develop a novel hybrid index structure, SAIR-tree, which comprehensively considers the attributes of social, spatial, and textual. Then, we propose an approximate algorithm and an exact algorithm, combining with the pruning strategy, can efficiently search the top-k result set. Finally, experiments on real dataset confirm the efficiency and accuracy of the proposed algorithms.