Bobed, C. (Universidad de Zaragoza) ; Bernad, J. (Universidad de Zaragoza) ; Ilarri, S. (Universidad de Zaragoza) ; Mena, E. (Universidad de Zaragoza)
Resumen: Approaches for the processing of location-dependent queries usually assume that the location data are expressed precisely, usually using GPS locations. However, this is unrealistic because positioning methods do not have a perfect accuracy (e.g., the positioning approach used in cellular networks handles only the cell where mobile users are located). Besides, users may need to express queries based on concepts of locations other than traditional GPS locations, which we call location granules. In this paper, we focus on location granule-based query processing (i.e., processing of queries with location granules) in situations where the location data available is imprecise, which we have called probabilistic location-dependent queries. For that purpose, we exploit the concept of uncertainty location granule, which represents the location uncertainty of an object. In particular, we tackle the problem of processing probabilistic inside (range) constraints. We analyze in detail how those constraints can be processed, taking into account both the existence of location uncertainty affecting the relevant objects and the location granularity specified. An extensive experimental evaluation shows the feasibility of the proposed probabilistic query processing approach and analyzes the advantages of using index structures to speed up the query processing. Idioma: Inglés DOI: 10.1016/j.pmcj.2016.12.006 Año: 2017 Publicado en: PERVASIVE AND MOBILE COMPUTING 39 (2017), 304-321 ISSN: 1574-1192 Factor impacto JCR: 2.974 (2017) Categ. JCR: COMPUTER SCIENCE, INFORMATION SYSTEMS rank: 33 / 148 = 0.223 (2017) - Q1 - T1 Categ. JCR: TELECOMMUNICATIONS rank: 27 / 87 = 0.31 (2017) - Q2 - T1 Factor impacto SCIMAGO: 0.368 - Computer Science (miscellaneous) (Q2) - Applied Mathematics (Q3)