Search

With the proliferation of smartphone cameras and social networks, we have rich, multi-modal data about points of interest (POIs) - like cultural landmarks, institutions, businesses, etc. - within a given areas of interest (AOI) (e.g., a county, city or a neighborhood) available to us. Data conflation across multiple modalities of data sources is one of the key challenges in maintaining a geographical information system (GIS) which accumulate data about POIs. Given POI data from nine different sources, and ground-level geo-tagged and scene-captioned images from two different image hosting platforms, in this work we explore the application of graph neural networks (GNNs) to perform data conflation, while leveraging a natural graph structure evident in geospatial data. The preliminary results demonstrate the capacity of a GNN operation to learn distributions of entity (POIs and images) features, coupled with topological structure of entity's local neighborhood in a semantic nearest neighbor graph, in order to predict links between a pair of entities.

Relief and recovery from disasters (both natural and human-made) require a coordinated approach across several federal and state government agencies. In order to achieve optimal resource allocation and deployment of first responders, accurate and timely assessment of the impact and extent of destruction are the cornerstones to any recovery effort. Ideally, this knowledge should be gathered and shared within the first 0-24 hours (termed as "Acute Phase" by the U.S. CDC guideline) for informed decision-making. But achieving this poses significant challenges for the data collection and data harmonization processes, particularly when voluminous data are being generated from diverse and distributed sources during the disaster responses. To this end, this work developed a scalable and efficient workflow to dynamically collect and harmonize crowd-sourced geographic multi-modal data, and then assess critical infrastructure (CI) damaged during disaster events. We demonstrate the application of our framework with two real-world experiences in addressing post-disaster recovery efforts - for the Bahamas (Natural - due to Hurricane Dorian, 2019) and Beirut (Human-made - due to explosion caused by the ammonium nitrate stored in a warehouse, 2020). We have illustrated that a coordinated effort is needed for planning as well as for execution to achieve informed decision making.

The opioid crisis has hit American cities hard, and research on spatial and temporal patterns of drug-related activities including detecting and predicting clusters of crime incidents involving particular types of drugs is useful for distinguishing hot zones where drugs are present that in turn can further provide a basis for assessing and providing related treatment services. In this study, we investigated spatiotemporal patterns of more than 52,000 reported incidents of drug-related crime at block group granularity in Chicago, IL between 2016 and 2019. We applied a space-time analysis framework and machine learning approaches to build a model using training data that identified whether certain locations and built environment and sociodemographic factors were correlated with drug-related crime incident patterns, and establish the top contributing factors that underlaid the trends. Space and time, together with multiple driving factors, were incorporated into a random forest model to analyze these changing patterns. We accommodated both spatial and temporal autocorrelation in the model learning process to assist with capturing the changes over time and tested the capabilities of the space-time random forest model by predicting drug-related activity hot zones. We focused particularly on crime incidents that involved heroin and synthetic drugs as these have been key drug types that have highly impacted cities during the opioid crisis in the U.S.

Conventional approaches for modeling human mobility pattern often focus on human activity and movement dynamics in their regular daily lives and cannot capture changes in human movement dynamics in response to large-scale events. With the rapid advancement of information and communication technologies, many researchers have adopted alternative data sources (e.g., cell phone records, GPS trajectory data) from private data vendors to study human movement dynamics in response to large-scale natural or societal events. Big geosocial data such as georeferenced tweets are publicly available and dynamically evolving as real-world events are happening, making it more likely to capture the real-time sentiments and responses of populations. However, precisely-geolocated geosocial data is scarce and biased toward urban population centers. In this research, we developed a big geosocial data analytical framework for extracting human movement dynamics in response to large-scale events from publicly available georeferenced tweets. The framework includes a two-stage data collection module that collects data in a more targeted fashion in order to mitigate the data scarcity issue of georeferenced tweets; in addition, a variable bandwidth kernel density estimation(VB-KDE) approach was adopted to fuse georeference information at different spatial scales, further augmenting the signals of human movement dynamics contained in georeferenced tweets. To correct for the sampling bias of georeferenced tweets, we adjusted the number of tweets for different spatial units (e.g., county, state) by population. To demonstrate the performance of the proposed analytic framework, we chose an astronomical event that occurred nationwide across the United States, i.e., the 2017 Great American Eclipse, as an example event and studied the human movement dynamics in response to this event. However, this analytic framework can easily be applied to other types of large-scale events such as hurricanes or earthquakes.

It is of interest to federal and state agencies to develop an advanced uniform method for estimation of vehicle miles traveled (VMT) on local roads which can be used as a guideline for agencies nationwide. The purpose of this study is to propose advanced innovative approaches for estimating VMT on local roads and analyze the feasibility of applying these methods. The paper presents a methodology and procedure for estimating local road VMT using GPS vehicle trajectory data and an all-street road network and expands these methodologies and results to determine the minimum required GPS sample size. The Federal Highway Administration and other transportation agencies may consider using these methodologies as a future guide to update VMT estimates with minimal additional cost requirements. The key finding of the research is that it is feasible to use new GPS vehicle trajectory data to estimate VMT on non-Federal Aid System roadways. The sample size of this data allows the application of this new method across the nation. The accuracy of this method was tested for the State of Maryland. Once such statewide GPS data is obtained by a given state, the methodology can be easily applied to that state as well.

Pedal misapplications by drivers have received attention as being an underlying factor for the phenomenon known as sudden unintended acceleration (SUA) in vehicles. This research investigates behaviors during a common task for drivers, namely residential parking. Parking has been identified as a maneuver that is often linked with SUA mishaps. Using driving trajectories data from a set of four couples collected as part of a naturalistic driving study, we investigate whether consistent behaviors can be detected when parking at home from a geospatial perspective, i.e., whether deceleration and braking occur in a characteristic way at the end of a driving trajectory, and whether these behaviors vary when the geospatial context of parking changes. An ontology-based approach is used to frame the key behaviors of the naturalistic driving, and big data techniques are applied to extract parking-specific behaviors from driving trajectories. Results show that individuals showed relatively consistent parking behaviors under the same geospatial context and the standard deviation of the deceleration threshold has a larger discrepancy between couples parking at different residences than within couples where parking occurs at the same place.

As location-sensing devices and apps become more prevalent, the scale and availability of big GPS trajectory data are also rapidly expanding. Big GPS trajectory data analytics offers new opportunities for gaining insights into vehicle movement dynamics and road network usage patterns that are important for transportation studies and urban planning among other fields. Processing big GPS trajectory data, consisting of billions of GPS waypoints and millions of individual trajectories is a challenging yet important task for researchers from these different domains. In this research, we propose an Apache Spark-based geo-computing framework for using big GPS trajectory data to estimate vehicle miles travelled, an important metric used by both federal and state highway agencies in the United States for transportation planning. The computing challenge lies in scaling the processing of billions of raw GPS points data as well as the steps for map matching for a statewide road network consisting of thousands of road segments. In this work, we develop a scalable map-matching module that considers both the spatiotemporal information of GPS waypoint sequences and topologic information of road network for the State of Maryland while striking a balance between matching accuracy and computing time. We processed 19.8 million raw GPS trips consisting of approximately 1.4 billion GPS waypoints collected in Maryland during a four-month period in 2015 to estimate vehicle miles travelled for Maryland’s road network. The estimation results show that using big GPS trajectory analytic methods is promising for obtaining accurate and stable vehicle miles travelled estimates.

In this chapter, we present a semantic data modeling framework for representing and analyzing the movement dynamics of individuals that arise from following a schedule or plan of activities in a semantic-enriched environment, i.e., an environment for which an ontology-driven space-time activity knowledgebase has been constructed. The ontology-driven knowledgebase contains spatial, temporal, and semantic information about geospatial entities in the environment. The relations between geospatial entities in the environment are captured in the knowledgebase through the underlying ontology support. Movement by individuals on a university campus according to a semester-based course schedule is employed as a use case to demonstrate this framework. This work demonstrates an RDF-based semantic data model that is used for reasoning about movement, including the movement trajectories of students on campus based on weekly course schedules. Road network information is incorporated to generate movement trajectories more realistically and data from a campus course information system allows us to query and analyze aggregated movement dynamics on campus. This chapter discusses the advantages of a semantic data modeling approach over traditional data models for human activity and movement including the capability to incorporate different data sources into the analysis, and generate geographic visualizations of movement paths as well as the overall movement dynamics for a campus.

The Wikipedia database is a data source of immense richness and variety. Included in this database are thousands of geotagged articles, including, for example, almost real-time updates on current and historic natural hazards. This includes usercontributed information about the location of natural hazards, the extent of the disasters, and many details relating to response, impact, and recovery. In this research, a computational framework is proposed to detect spatial patterns of natural hazards from the Wikipedia database by combining topic modeling methods with spatial analysis techniques. The computation is performed on the Neon Cluster, a high performance-computing cluster at the University of Iowa. This work uses wildfires as the exemplar hazard, but this framework is easily generalizable to other types of hazards, such as hurricanes or flooding. Latent Dirichlet Allocation (LDA) modeling is first employed to train the entire English Wikipedia dump, transforming the database dump into a 500-dimension topic model. Over 230,000 geo-tagged articles are then extracted from the Wikipedia database, spatially covering the contiguous United States. The geo-tagged articles are converted into an LDA topic space based on the topic model, with each article being represented as a weighted multidimension topic vector. By treating each article’s topic vector as an observed point in geographic space, a probability surface is calculated for each of the topics. In this work, Wikipedia articles about wildfires are extracted from the Wikipedia database, forming a wildfire corpus and creating a basis for the topic vector analysis. The spatial distribution of wildfire outbreaks in the US is estimated by calculating the weighted sum of the topic probability surfaces using a map algebra approach, and mapped using GIS. To provide an evaluation of the approach, the estimation is compared to wildfire hazard potential maps created by the USDA Forest service.

In this paper, a spatiotemporal framework is developed for identifying building vulnerabilities and content evacuations during riverine flooding events. This work investigates the spatiotemporal properties required to trigger building contents evacuations in the floodplain during a flood event. The spatial properties for building risks are based on topography, flood inundation, building location, building elevation, and road access to determine five categories of vulnerability, vulnerable basement , flooded basement , vulnerable first-floor , flooded first - floor , and road access . Using this framework, a model designed to track the spatiotemporal patterns of building evacuations is presented. The model is based upon real-time flood forecast predictions that are linked with building properties to create a model that captures the spatiotemporal ordering of building vulnerabilities and building content evacuations. Applicable to different communities at risk from flooding, the evacuation model is applied to a historical flood for a university campus, demonstrating how the defined elements are used to derive a pattern of vulnerability and evacuation for a campus threatened by severe flooding.

Although influenza is a common disease with characteristics of seasonality, the determinants of each season's onset, magnitude, and duration are poorly understood. This paper focuses on the role of environmental factors in spread and epidemic of seasonal influenza and explores the environmental explanatory factors for different types of influenza in mainland China. It also introduces satellite remote sensing as an important data obtaining approach, and highlights the potential of using satellite images for monitoring dynamics of climate and landscapes related to the spread of seasonal influenza. Applying Geographic Information System (GIS) technique combining with traditional statistical analysis, the paper uses influenza virus isolation rate (VIR) as the measure of influenza activity and analyzed its association with environmental factors. The results show that the spread and epidemic of influenza is influenced by various environmental factors, among which the temperature and humidity are seemed to be the determinants. However, the impacts of the environmental factors to different types of virus are varied. Low temperature and humidity conditions arere associated with a higher activity of both influenza A and B. On the contrary, high temperature and humidity conditions are associated with a higher activity of influenza A, but are associated with only a moderate or low, less consistent increase in the activity of influenza B. Recognition of this association could lead to better understanding of the mechanisms of emergence of influence epidemics and provide scientific evidence for controlling influenza.

The non-pharmaceutical intervention to reduce the impact and spread of COVID-19 requires the development of policies and guidance through a collaborative effort among government, academia, medicine, and citizens. To operationalize this effort, we have developed an all-encompassing situational awareness platform that can process multi-modal and multi-source data allowing informed decision making. Besides, showing the current spread of infection, the platform also captures the impact of human dynamics on the infection spread, location, and availability of critical infrastructure, prediction, and high-performance computing driven simulation. The platform is extensible, allowing third-party integration and services to consume the curated data and analytics in near real-time. We believe the platform will augment critical decision making for reducing the impact and spread of the pandemic.