Xun Zhou

Experiences with evacuation route planning algorithms

Efficient tools are needed to identify routes and schedules to evacuate affected populations to safety in the event of natural disasters. Hurricane Rita and the recent tsunami revealed limitations of traditional approaches to provide emergency preparedness for evacuees and to predict the effects of evacuation route planning (ERP). Challenges arise during evacuations due to the spread of people over space and time and the multiple paths that can be taken to reach them; key assumptions such as stationary ranking of alternative routes and optimal substructure are violated in such situations. Algorithms for ERP were first developed by researchers in operations research and transportation science. However, these proved to have high computational complexity and did not scale well to large problems. Over the last decade, we developed a different approach, namely the Capacity Constrained Route Planner (CCRP), which generalizes shortest path algorithms by honoring capacity constraints and the spread of people over space and time. The CCRP uses time-aggregated graphs to reduce storage overhead and increase computational efficiency. Experimental evaluation and field use in Twin Cities Homeland Security scenarios demonstrated that CCRP is faster, more scalable, and easier to use than previous techniques. We also propose a novel scalable algorithm that exploits the spatial structure of transportation networks to accelerate routing algorithms for large network datasets. We evaluated our new approach for large-scale networks around downtown Minneapolis and riverside areas. This article summarizes experiences and lessons learned during the last decade in ERP and relates these to Professor Goodchilds contributions.

Spatiotemporal Data Mining: A Computational Perspective

Explosive growth in geospatial and temporal data as well as the emergence of new technologies emphasize the need for automated discovery of spatiotemporal knowledge. Spatiotemporal data mining studies the process of discovering interesting and previously unknown, but potentially useful patterns from large spatiotemporal databases. It has broad application domains including ecology and environmental management, public safety, transportation, earth science, epidemiology, and climatology. The complexity of spatiotemporal data and intrinsic relationships limits the usefulness of conventional data science techniques for extracting spatiotemporal patterns. In this survey, we review recent computational techniques and tools in spatiotemporal data mining, focusing on several major pattern families: spatiotemporal outlier, spatiotemporal coupling and tele-coupling, spatiotemporal prediction, spatiotemporal partitioning and summarization, spatiotemporal hotspots, and change detection. Compared with other surveys in the literature, this paper emphasizes the statistical foundations of spatiotemporal data mining and provides comprehensive coverage of computational approaches for various pattern families. ISPRS Int. J. Geo-Inf. 2015, 4 2307 We also list popular software tools for spatiotemporal data analysis. The survey concludes with a look at future research needs.

Discovering interesting sub-paths in spatiotemporal datasets: a summary of results

Given a spatiotemporal (ST) dataset and a path in its embedding spatiotemporal framework, the goal is to to identify all interesting sub-paths defined by an interest measure. Sub-path discovery is of fundamental importance for understanding climate changes, agriculture, and many other application. However, this problem is computationally challenging due to the massive volume of data, the varying length of sub-paths and non-monotonicity of interestingness throughout a sub-path. Previous approaches find interesting unit sub-paths (e.g., unit time interval) or interesting points. By contrast, we propose a Sub-path Enumeration and Pruning (SEP) approach that finds collections of long interesting sub-paths. Two case studies using climate change datasets show that SEP can find long interesting sub-paths which represent abrupt climate change. We provide theoretical analyses of correctness, completeness and computational complexity of the proposed approach. We also provide experimental evaluation of two traversal strategies for enumerating and pruning candidate sub-paths.

Spatiotemporal change footprint pattern discovery: an inter-disciplinary survey

Given a definition of change and a dataset about spatiotemporal (ST) phenomena, ST change footprint discovery is the process of identifying the location and/or time of such changes from the dataset. Change footprint discovery is fundamentally important for the study of climate change, the tracking of disease, and many other applications. Methods for detecting change footprints have emerged from a diverse set of research areas, ranging from time series analysis and remote sensing to spatial statistics. Researchers have much to learn from one another, but are stymied by inconsistent use of terminology and varied definitions of change across disciplines. Existing reviews focus on discovery methods for only one or a few types of change footprints (e.g., point change in a time series). To facilitate sharing of insights across disciplines, we conducted a multi‐disciplinary review of ST change patterns and their respective discovery methods. We developed a taxonomy of possible ST change footprints and classified our review findings accordingly. This exercise allowed us to identify gaps in the research that we consider ripe for exploration, most notably change pattern discovery in vector ST datasets. In addition, we illustrate how such pattern discovery might proceed using two case studies from historical GIS. WIREs Data Mining Knowl Discov 2014, 4:1–23. doi: 10.1002/widm.1113

Focal-Test-Based Spatial Decision Tree Learning

Given learning samples from a raster data set, spatial decision tree learning aims to find a decision tree classifier that minimizes classification errors as well as salt-and-pepper noise. The problem has important societal applications such as land cover classification for natural resource management. However, the problem is challenging due to the fact that learning samples show spatial autocorrelation in class labels, instead of being independently identically distributed. Related work relies on local tests (i.e., testing feature information of a location) and cannot adequately model the spatial autocorrelation effect, resulting in salt-and-pepper noise. In contrast, we recently proposed a focal-test-based spatial decision tree (FTSDT), in which the tree traversal direction of a sample is based on both local and focal (neighborhood) information. Preliminary results showed that FTSDT reduces classification errors and salt-and-pepper noise. This paper extends our recent work by introducing a new focal test approach with adaptive neighborhoods that avoids over-smoothing in wedge-shaped areas. We also conduct computational refinement on the FTSDT training algorithm by reusing focal values across candidate thresholds. Theoretical analysis shows that the refined training algorithm is correct and more scalable. Experiment results on real world data sets show that new FTSDT with adaptive neighborhoods improves classification accuracy, and that our computational refinement significantly reduces training time.

Ring-Shaped Hotspot Detection: A Summary of Results

Given a collection of geo-located activities (e.g., Crime reports), ring-shaped hotspot detection (RHD) finds rings, where concentration of activities inside the ring is much higher than outside. RHD is important for the applications such as crime analysis, where it may focus the search for crime sources location, e.g. The home of a serial criminal. RHD is challenging because of the large number of candidate rings and the high computational cost of the statistical significance test. Previous statistically significant hotspot detection techniques (e.g., Sat Scan) identify circular/rectangular areas, but can not discover rings. This paper proposes a dual grid based pruning (DGP) approach to detect ring-shaped hotspots. A case study on real crime data confirms that DGP detects novel ring-shaped regions, regions that go undetected by Sat Scan. Experiments show that DGP improves the computational cost of a naive approach substantially.

Focal-Test-Based Spatial Decision Tree Learning: A Summary of Results

Given a raster spatial framework, as well as training and test sets, the spatial decision tree learning (SDTL) problem aims to minimize classification errors as well as salt-and-pepper noise. The SDTL problem is important due to many societal applications such as land cover classification in remote sensing. However, the SDTL problem is challenging due to the spatial autocorrelation of class labels, and the potentially exponential number of candidate trees. Related work is limited due to the use of local-test-based decision nodes, which can not adequately model spatial autocorrelation during test phase, leading to high salt-and-pepper noise. In contrast, we propose a focal-test-based spatial decision tree (FTSDT) model, where the tree traversal direction for a location is based on not only local but also focal (i.e., neighborhood) properties of the location. Experimental results on real world remote sensing datasets show that the proposed approach reduces salt-and-pepper noise and improves classification accuracy.

FAST: a generic framework for flash-aware spatial trees

Spatial tree index structures are crucial components in spatial data management systems, designed with the implicit assumption that the underlying external memory storage is the conventional magnetic hard disk drives. This assumption is going to be invalid soon, as flash memory storage is increasingly adopted as the main storage media in mobile devices, digital cameras, embedded sensors, and notebooks. Though it is direct and simple to port existing spatial tree index structures on the flash memory storage, that direct approach does not consider the unique characteristics of flash memory, i.e., slow write operations, and erase-before-update property, which would result in a sub optimal performance. In this paper, we introduce FAST (i.e., Flash-Aware Spatial Trees) as a generic framework for flash-aware spatial tree index structures. FAST distinguishes itself from all previous attempts of flash memory indexing in two aspects: (1) FAST is a generic framework that can be applied to a wide class of data partitioning spatial tree structures including R-tree and its variants, and (2) FAST achieves both efficiency and durability of read and write flash operations through smart memory flushing and crash recovery techniques. Extensive experimental results, based on an actual implementation of FAST inside the GiST index structure in PostgreSQL, show that FAST achieves better performance than its competitors.

Significant Route Discovery: A Summary of Results

Given a spatial network and a collection of activities (e.g., pedestrian fatality reports, crime reports), Significant Route Discovery (SRD) finds all shortest paths in the spatial network where the concentration of activities is unusually high (i.e., statistically significant). SRD is important for societal applications in transportation safety, public safety, or public health such as finding routes with significant concentrations of accidents, crimes, or diseases. SRD is challenging because 1) there are a potentially large number of candidate routes (~1016) in a given dataset with millions of activities or road network nodes and 2) significance testing does not obey the monotonicity property. Previous work focused on finding circular areas of concentration, limiting its usefulness for finding significant linear routes on a network. SaTScan may miss many significant routes since a large fraction of the area bounded by circles for activities on a path will be empty. This paper proposes a novel algorithm for discovering statistically significant routes. To improve performance, the proposed algorithm features algorithmic refinements that prune unlikely paths and speeds up Monte Carlo simulation. We present a case study comparing the proposed statistically significant network-based analysis (i.e., shortest paths) to a statistically significant geometry-based analysis (e.g., circles) on pedestrian fatality data. Experimental results on real data show that the proposed algorithm, with our algorithmic refinements, yields substantial computational savings without reducing result quality.

A vision for GPU-accelerated parallel computation on geo-spatial datasets

We summarize the need and present our vision for accelerating geo-spatial computations and analytics using a combination of shared and distributed memory parallel platforms, with general-purpose Graphics Processing Units (GPUs) with 100s to 1000s of processing cores in a single chip forming a key architecture to parallelize over. A GPU can yield one-to-two orders of magnitude speedups and will become increasingly more affordable and energy efficient due to mass marketing for gaming. We also survey the current landscape of representative geo-spatial problems and their parallel, GPU-based solutions.