On Efficiently Diversified Top-K Geo-Social Keyword Query Processing in Road Networks

Jingwen Zhao,Yunjun Gao,Chunyu Ma,Pengfei Jin,Shiting Wen
DOI: https://doi.org/10.1016/j.ins.2019.10.021
IF: 8.1
2019-01-01
Information Sciences
Abstract:With the development of geo-positioning technologies, the volumes of spatio-textual data are growing rapidly. These data are available in many applications such as location-based services and geo-social networks, and thus, various types of geo-social keyword queries have been intensively studied in the literature. A top-k geo-social keyword (TkGSK) query retrieves the objects by considering spatial, social, and textual constraints between the query and objects. Moreover, the result diversification is becoming a common practice to enhance the quality of the query result. Motivated by these, in this paper, we study the problem of diversified top-k geo-social keyword (DkGSK) query, which considers not only the relevance but also the diversity of the result. We first prove that this problem is NP-hard, and then, we present an exact algorithm with several effective pruning strategies. Also, we develop an approximate algorithm with proved approximation ratio to support the DkGSK query. Considerable experiments on real data sets demonstrate the effectiveness and efficiency of our proposed algorithms.
What problem does this paper attempt to address?