Cooperative treatment of failing queries over uncertain databases: a matrix-computation-based approach

摘要

A large number of applications such as sensor networks, RFID-based monitoring systems, mobile object management and location-based services manage data pervaded with uncertainty. Usually, users wish/prefer high quality results (i.e. with highest certainty) when they pose queries with strict conditions over these data. However, as they may not be clear about the contents of the databases that contain such data, these queries may be failing i.e., they may return no result or results that do not satisfy the expected level of certainty. In this case, users may try to change manually the query conditions to obtain approximate answers. Due to the exponential combination number of query conditions, this procedure results in a time-consuming and frustrating task. In this paper, we address the failing queries problem by proposing an approach that identifies the query parts, called Minimal Failing Subqueries (MFSs), that are responsible for its failure. Thanks to these MFSs, interactive and automatic approaches can be set up to help the user reformulating her/his query. We also compute, in the same time, a set of Maximal Succeeding Subqueries (XSSs) that represents a list of non failing queries with a maximal number of predicates of the initial query. The results of these XSSs constitute good alternative answers that can be returned to the user instead of an empty result. To demonstrate the efficiency and the effectiveness of our proposal, a set of experiments have been conducted with synthetic and real datasets. A comparison with baseline and related work approaches shows the interest of our proposal.