Uncertain Spatial Data Management:An Overview

Abstract

Both the current trends in technology such as smart phones, general mobile devices, stationary sensors, and satellites as we as a new user mentality of using this technology to voluntarily share enriched location information produces a flood of geo-spatial and geo-spatio-temporal data. This data flood provides a tremendous potential of discovering new and useful knowledge. But in addition to the fact that measurements are imprecise, spatial data is often interpolated between discrete observations. To reduce communication and bandwidth utilization, data is often subjected to a reduction, thereby eliminating some of the known/recorded values. These issues introduce the notion of uncertainty in the context of spatio-temporal data management, an aspect raising imminent need for scalable and flexible solutions. The main scope of this chapter is to survey existing techniques for managing, querying, and mining uncertain spatio-temporal data. First, this chapter surveys common data representations for uncertain data, explains the commonly used possible worlds semantics to interpret an uncertain database, and surveys existing system to process uncertain data. Then this chapter defines the notion of different probabilistic result semantics to distinguish the task of enrich individual objects with probabilities rather than enriched entire results with probabilities. To distinguish between result semantics is important, as for many queries, the problem of computing object-level result probabilities can be done efficiently, whereas the problem of computing probabilities of entire results is often exponentially hard. Then, this chapter provides an overview over probabilistic query predicates to quantify the required probability of a result to be included in the result. Finally, this chapter introduces a novel paradigm to efficiently answer any kind of query on uncertain data: the Paradigm of Equivalent Worlds, which groups the exponential set of possible database worlds into a polynomial number of set of equivalent worlds that can be processed efficiently. Examples and use-cases of querying uncertain spatial data are provided using the example of uncertain range queries.

Spatial Data Quality in the Internet of Things: Management, Exploitation, and Prospects

“…In this setting, query processing techniques estimate upper and lower bounds of query objects based on probability models to enable priority-oriented processing and object pruning. A taxonomy of probabilistic spatial queries is available [53], and a recent survey [284] categorizes the existing queries over uncertain spatial data according to query types. In contrast, we categorize query processing techniques based on the type of location uncertainty they handle in the context of IoT-based localization/tracking, namely the uncertainty caused by inaccuracy of localization algorithms and that caused by the discrete sampling of devices [176].…”

“…Queries over uncertain spatial data have been studied extensively in the last decades, while how to query uncertain SID in a resource-limited and stream setting remains open [284].…”

“…On the other hand, positioning technologies enabled by the wireless communication infrastructure and wireless and ambient devices have been integrated widely into the IoT infrastructure (called Location of Things [186]) to provide spatial references to other IoT devices. We call such spatially [117] -✓ ✓ characteristics of big geospatial data and issues in handling them Goodchild [70,71] -✓ ✓ quality and uncertainty issues of big geodata processing Guptill and Morrison [76] -✓ ✓ elements of evaluating spatial data quality Devillers et al [58] -✓ ✓ recent achievement and open issues in improving spatial data quality Li et al [108] -✓ ✓ quality assessment tools and uncertainty-aware spatial analytics Züfle et al [284,285] -✓ ✓ major challenges in handling uncertain geospatial data Zheng and Su [279] -✓ ✓ quality and semantics of raw trajectory data Tsai et al [208] ✓ • -features of IoT data and data mining techniques for IoT Siow et al [187] ✓ • -IoT and big data analytics in creating applications and services Mohammadi et al [158] ✓ • -deep learning for IoT big data and stream data analytics Karkouch et al [95] ✓ ✓ -IoT factors endangering data quality and IoT outlier detection Banerjee et al [24] ✓ ✓ -human-in-the-loop for IoT data quality control Ann and Wagh [16] ✓ ✓ -IoT data testing layer for data quality assurance Perez-Castillo et al [175] ✓ ✓ -IoT data quality in smart, connected product (SCP) environments Liu et al [131] ✓ ✓ -IoT data quality dimensions and related measurement methods Song and Zhang [189] ✓ ✓ -deep learning for validity, completeness, and consistency of IoT data Shit et al [186] ✓ -✓ analysis and taxonomy of IoT-based localization techniques Javarneh et al [12] ✓ -✓ cloud-based big spatial data management frameworks for IoT Mahdavinejad et al [152] ✓ • • machine learning methods for IoT smart cities Chen et al [47] ✓ • ✓ robustness, security, and privacy of IoT-enabled Location-based Services Li et al [122] ✓ • ✓ error sources and mitigation methods of IoT-signal-based localization Ours ✓ ✓ ✓ quality management of SID and exploitation of low-quality SID ✓ focused, • partially covered, -not mentioned referenced data from IoT devices spatial IoT data (SID). Two important special cases of SID are distinguished: trajectories, as time series of location values; and spatiotempora...…”

Spatial Data Quality in the IoT Era: Management and Exploitation

“…In this setting, algorithms estimate upper and lower bounds of query objects based on probability models to enable priorityoriented processing and object pruning. A taxonomy of probabilistic spatial queries is available [27], and a recent survey [140] categorizes queries over uncertain spatial data. In contrast, the tutorial presents query processing techniques based on the type of location uncertainty they handle in the context of IoT-based positioning or tracking: First, to handle uncertainty caused by location inaccuracy, an object's location at a single time is usually described as a probability density function (pdf), which occurs in continuous (a closed-form distribution) [12,13,26,68,100] or discrete (a set of samples with occurrence probabilities) cases [43,120,131].…”

“…The popularity of the studies on SID quality issues is also evidenced by an increasing emergence of survey papers. However, most of existing survey papers focus on synergies between two of the three related areas, i.e., the IoT, data quality, and spatial computing, covering topics such as IoT data quality [11,50,71,96], spatial data quality [28,35,38,135,140,141], and IoT-enabled spatial applications [5,80,94]. Although several survey papers [66,80] on IoTenabled spatial applications mention quality issues, they do not analyze and summarize DQ technologies.…”