Fine-scale soil mapping in Alaska, traditionally relying on fieldwork and localized simulations, remains a critical yet underdeveloped task, despite the region's ecological importance and extensive permafrost coverage. As permafrost thaw accelerates due to climate change, it threatens infrastructure stability and key ecosystem services, such as soil carbon storage. High-resolution soil maps are essential for characterizing permafrost and soil taxonomy distributions, identifying vulnerable areas, and informing adaptation strategies. In this paper, we present MiSo, a machine learning (ML) model to produce statewide fine-scale soil maps for near-surface permafrost and soil taxonomy. MiSo integrates a geospatial foundation model for visual feature extraction, implicit neural representations for continuous spatial prediction, and contrastive learning for multimodal alignment and geo-location awareness. We compare MiSo with Random Forest (RF), a traditional ML model that has been widely used in soil mapping applications. Spatial cross-validation and regional analysis across Permafrost Zones and Major Land Resource Areas (MLRAs) show that MiSo generalizes better to remote, unseen locations and achieves higher recall than RF, which is critical for monitoring permafrost thaw and related environmental processes. These findings demonstrate the potential of advanced ML approaches for fine-scale soil mapping and provide practical guidance for future soil sampling and infrastructure planning in permafrost-affected landscapes.

Predicting human mobility across multiple cities is essential for urban applications but remains challenging due to the complex and diverse spatiotemporal dynamics in human trajectories. While recent work often leverages language modeling by treating trajectories as sequences for next-location prediction, these approaches typically rely on raw movement data, process long trajectories without distinguishing between individual trips, and use a single model for all users within a city. To address these limitations, this paper presents MoVER, a transformer encoder-decoder that enriches trajectory representations with location profiles and explicit trip separators. Furthermore, we introduce a clustering-based finetuning strategy to handle user heterogeneity by tailoring models to user groups of similar travel patterns. MoVER outperforms baselines on a validation set and achieves a top-7 ranking among over 50 participating teams in the 14th SIGSPATIAL Cup competition (GISCUP 2025). The code is publicly available at: https://github.com/knowledge-computing/giscup2025-mover.

Geospatial question answering (QA) is a fundamental task in navigation and point of interest (POI) searches, yet existing datasets are limited in scale, diversity, and they rely on text-only descriptions without incorporating geometries. We introduce MapQA, a dataset that couples question-answer pairs with geo-entity geometries from OpenStreetMap (OSM) across two regions (Southern California and Illinois). MapQA contains 3,154 QA pairs covering nine geospatial reasoning types, including neighborhood inference and type identification, expanding both the quantity and variety of existing resources. To evaluate methods, we compare (1) a retrieval-based model that ranks geo-entities by embedding similarity and (2) large language models (LLMs) that translate questions into SQL queries executed on OSM. Retrieval-based models capture spatial relations like closeness and direction but fail on explicit distance computations, while LLMs excel at one-hop reasoning yet struggle with multi-hop tasks, revealing a key challenge for future systems. MapQA is publicly available at https://github.com/knowledge-computing/MapQA-dataset.

Historical geologic maps contain rich geospatial information—such as rock units, faults, folds, and bedding planes—that is critical for assessing mineral resources essential to renewable energy, electric vehicles, and national security. However, digitizing maps remains a labor-intensive and time-consuming task. We present DIGMAPPER, a modular, scalable system developed in collaboration with the United States Geological Survey (USGS) to automate the digitization of geologic maps. DIGMAPPER features a fully dockerized, workflow-orchestrated architecture that integrates state-of-the-art deep learning models for map layout analysis, feature extraction, and georeferencing. To overcome challenges such as limited training data and complex visual content, our system employs innovative techniques, including in-context learning with large language models, synthetic data generation, and transformer-based models. Evaluations on over 100 annotated maps from the DARPA-USGS dataset demonstrate high accuracy across polygon, line, and point feature extraction, and reliable georeferencing performance. Deployed at USGS, DIGMAPPER significantly accelerates the creation of analysis-ready geospatial datasets, supporting national-scale critical mineral assessments and broader geoscientific applications.

Historical maps often suffer from coloring errors caused by artifacts during map production or scanning. These errors result in color mismatches between important map features (e.g., polygon layers) and their corresponding map keys, which hinders both human interpretation and automated feature extraction. This paper targets the problem of automatically correcting polygon coloring errors in historical maps using only in-map information, such as the map keys. The challenge lies in the diverse visual representations of map keys and variations in coloring errors, which differ significantly both within and across maps. We propose a machine-learning model that automatically identifies and corrects color inconsistencies between map polygon layers and their visual appearances defined by the map keys on the same map. Our approach leverages polygon metadata, such as map keys describing the visual and semantic properties of each polygon on maps, to detect mismatches in color histograms and representations and recolor the incorrect areas in the map content accordingly. We evaluate our approach on USGS geological maps; it outperforms comparative methods by at least 7.5%. In addition, our approach improves the downstream automated polygon-extraction task by 18.0% in precision.

Black-and-white draft geological maps, produced during fieldwork, often contain dense handwritten annotations overlaid on monochromatic contour basemaps. Although interpretable in grayscale, the lack of color makes it difficult to visually distinguish overlapping or adjacent geological units, especially when boundaries are unclear and annotation styles vary. However, colorizing these draft maps is labor-intensive but essential, as they may be the only source of detailed geological information for certain regions. This hinders both human interpretation and downstream tasks such as map digitization and critical mineral resource assessment. We target the problem of automated colorization of draft geological maps. The challenge lies in interpreting noisy visual cues from uncolored sketches and assigning appropriate colors according to their geological categories. We propose a novel machine learning approach that exploits polygon metadata, including map keys that explicitly define geological units and implicitly suggest their intended colors, along with the semantic interpretation of the sketch content in the maps, to assign colors to the draft maps accordingly. We evaluate our method on USGS draft geological maps; it outperforms comparative methods by 15.7%. In addition, our approach improves downstream polygon-extraction performance by 9% in F1 score.

Ensuring equitable public transit access remains challenging, particularly in densely populated cities like New York City (NYC), where low-income and minority communities often face limited transit accessibility. Bike-sharing systems (BSS) can bridge these equity gaps by providing affordable first- and last-mile connections. However, strategically expanding BSS into underserved neighborhoods is difficult due to uncertain bike-sharing demand at newly planned ("cold-start") station locations and limitations in traditional accessibility metrics that may overlook realistic bike usage potential. We introduce Transit for All (TFA), a spatial computing framework designed to guide the equitable expansion of BSS through three components: (1) spatially-informed bike-sharing demand prediction at cold-start stations using region representation learning that integrates multimodal geospatial data, (2) comprehensive transit accessibility assessment leveraging our novel weighted Accessibility Level (wAL) by combining predicted bike-sharing demand with conventional transit accessibility metrics, and (3) strategic recommendations for new bike station placements that consider potential ridership and equity enhancement. Using NYC as a case study, we identify transit accessibility gaps that disproportionately impact low-income and minority communities in historically underserved neighborhoods. Our results show that strategically placing new stations guided by wAL notably reduces disparities in transit access related to economic and demographic factors. From our study, we demonstrate that TFA provides practical guidance for urban planners to promote equitable transit and enhance the quality of life in underserved urban communities.

Text on historical maps contains valuable information providing georeferenced historical, political, and cultural contexts. However, text extraction from historical maps has been challenging due to the lack of (1) effective methods and (2) training data. Previous approaches use ad-hoc steps tailored to only specific map styles. Recent machine learning-based text spotters (e.g., for scene images) have the potential to solve these challenges because of their flexibility in supporting various types of text instances. However, these methods remain challenges in extracting precise image features for predicting every sub-component (boundary points and characters) in a text instance. This is critical because map text can be lengthy and highly rotated with complex backgrounds, posing difficulties in detecting relevant image features from a rough text region. This paper proposes PALETTE, an end-to-end text spotter for scanned historical maps of a wide variety. PALETTE introduces a novel hyper-local sampling module to explicitly learn localized image features around the target boundary points and characters of a text instance for detection and recognition. PALETTE also enables hyper-local positional embeddings to learn spatial interactions between boundary points and characters within and across text instances. In addition, this paper presents a novel approach to automatically generate synthetic map images, SYNTHMAP+, for training text spotters for historical maps. The experiment shows that PALETTE with SYNTHMAP+ outperforms SOTA text spotters on two new benchmark datasets of historical maps, particularly for long and angled text. We have deployed PALETTE with SYNTHMAP+ to process over 60,000 maps in the David Rumsey Historical Map collection and generated over 100 million text labels to support map searching.

In various application domains like transportation, urban planning, and public health, analyzing human mobility, represented as a sequence of consecutive visits (aka trajectories), is crucial for uncovering essential mobility patterns. Current practices often discretize space and time to model trajectory data with sequence-analysis techniques like Transformers and LSTM, but this discretization tends to obscure the intrinsic spatial and temporal characteristics inherent in trajectories. Recent work shows the effectiveness of modeling trajectories directly in continuous space and time using the spatiotempo-ral point process (STPP). However, these approaches often assume that all observed trajectories originate from a single underlying dynamic. In reality, real-world trajectories exhibit varying dynamics or moving patterns. We hypothesize that grouping trajectories governed by similar dynamics into clusters before trajectory modeling could enhance modeling effectiveness. Thus, we present a novel approach that simultaneously models trajectories in continuous space and time using STPP while clustering them. Our method leverages a variational Expectation-Maximization (EM) framework to iteratively improve the learning of trajectory dynamics and refine cluster assignments within a single training phase. Extensive tests on synthetic and real-world data demonstrate its effectiveness in clustering and modeling trajectories.

Forecasting the number of visits to Points-of-Interest (POI) in an urban area is critical for planning and decision making in various application domains, from urban planning and transportation management to public health and social studies. Although this forecasting problem can be formulated as a multivariate time-series forecasting task, current approaches cannot fully exploit the ever-changing multi-context correlations among POIs. Therefore, we propose Busyness Graph Neural Network (BysGNN), a temporal graph neural network designed to learn and uncover the underlying multi-context correlations between POIs for accurate visit forecasting. Unlike other approaches where only time-series data is used to learn a dynamic graph, BysGNN utilizes all contextual information and time-series data to learn an accurate dynamic graph representation. By incorporating all contextual, temporal, and spatial signals, we observe a significant improvement in our forecasting accuracy over state-of-the-art forecasting models in our experiments with real-world datasets across the United States.

Scanned historical maps in libraries and archives are valuable repositories of geographic data that often do not exist elsewhere. Despite the potential of machine learning tools like the Google Vision APIs for automatically transcribing text from these maps into machine-readable formats, they do not work well with large-sized images (e.g., high-resolution scanned documents), cannot infer the relation between the recognized text and other datasets, and are challenging to integrate with post-processing tools. This paper introduces the mapKurator system, an end-to-end system integrating machine learning models with a comprehensive data processing pipeline. mapKurator empowers automated extraction, post-processing, and linkage of text labels from large numbers of large-dimension historical map scans. The output data, comprising bounding polygons and recognized text, is in the standard GeoJSON format, making it easily modifiable within Geographic Information Systems (GIS). The proposed system allows users to quickly generate valuable data from large numbers of historical maps for in-depth analysis of the map content and, in turn, encourages map findability, accessibility, interoperability, and reusability (FAIR principles). We deployed the mapKurator system and enabled the processing of over 60,000 maps and over 100 million text/place names in the David Rumsey Historical Map collection. We also demonstrated a seamless integration of mapKurator with a collaborative web platform to enable accessing automated approaches for extracting and linking text labels from historical map scans and collective work to improve the results.

The Spatial Triad model provides a framework for studying human interactions and experiences with the environment, which helps to improve human well-being and quality of life. Typical studies that use this framework require time-consuming and expensive surveys. This paper presents a simple yet effective approach to learning what humans feel, think, and see about their surroundings from easily accessible online text descriptions (e.g., descriptions of listings on real estate or travel blogs). The proposed technologies learn meaningful document and locality representations in a unified representation space, capturing important concepts shared among documents within the same locality. The proposed approach outperforms the existing method in finding associated localities in online text and shows exciting insights into locality similarity using the learned representations.

Recent advances in incorporating physical knowledge into deep neural networks can estimate previously unknown governing partial differential equations (PDEs) in a data-driven way. They have shown promising results in spatiotemporal predictive learning. However, these methods typically assume universal governing PDEs across space, which is impractical for modeling complex spatiotemporal phenomena with high spatial variability (e.g., climate). Also, they cannot effectively model the evolution of potential errors in estimating the physical dynamics over time. This paper introduces a physics-guided neural network, SVPNet, which learns effective physical representations by estimating the error evolution in physics states for correction and modeling spatially varying physical dynamics to predict the next state. Experiments carried out in four scenarios, including benchmarks and real-world datasets, show that SVPNet outperforms state-of-the-art methods in spatiotemporal prediction tasks for natural processes and significantly improves prediction when training data are limited. Ablation studies also highlight that SVPNet is powerful in capturing physical dynamics in complex physical systems.

Locating undiscovered deposits of critical minerals requires accurate geological data. However, most of the 100,000 historical geological maps of the United States Geological Survey (USGS) are in raster format. This hinders critical mineral assessment. We target the problem of extracting geological features represented as polygons from raster maps. We exploit the polygon metadata that provides information on the geological features, such as the map keys indicating how the polygon features are represented, to extract the features. We present a metadata-driven machine-learning approach that encodes the raster map and map key into a series of bitmaps and uses a convolutional model to learn to recognize the polygon features. We evaluated our approach on USGS geological maps; our approach achieves a median F1 score of 0.809 and outperforms state-of-the-art methods by 4.52%.

Accessing realistic human movements (aka trajectories) is essential for many application domains, such as urban planning, transportation, and public health. However, due to privacy and commercial concerns, real-world trajectories are not readily available, giving rise to an important research area of generating synthetic but realistic trajectories. Inspired by the success of deep neural networks (DNN), data-driven methods learn the underlying human decision-making mechanisms and generate synthetic trajectories by directly fitting real-world data. However, these DNN-based approaches do not exploit people's moving behaviors (e.g., work commute, shopping purpose), significantly influencing human decisions during the generation process. This paper proposes MBP-GAIL, a novel framework based on generative adversarial imitation learning that synthesizes realistic trajectories that preserve moving behavior patterns in real data. MBP-GAIL models temporal dependencies by Recurrent Neural Networks (RNN) and combines the stochastic constraints from moving behavior patterns and spatial constraints in the learning process. Through comprehensive experiments, we demonstrate that MBP-GAIL outperforms state-of-the-art methods and can better support decision making in trajectory simulations.

Changes in place names offer insight into regions' culture, politics, and geographical characteristics. This paper proposes an automatic approach to generate time-sequenced maps that show place name changes on a large number of maps from different time periods. The proposed approach utilizes gazetteers (i.e., indexes for geographic names) to retrieve a place's coordinates and name variants and searches for text labels from maps matching those coordinates and names. The resulting maps give rich, visual insights into how place names change over time and could facilitate historians' investigation of geographic name changes at a large scale.

Human mobility clustering is an important problem for understanding human mobility behaviors (e.g., work and school commutes). Existing methods typically contain two steps: choosing/learning a mobility representation and applying a clustering algorithm to the representation. However, these methods rely on strict visiting orders in trajectories and cannot take advantage of multiple types of mobility representations. This paper proposes a novel mobility clustering method for mobility behavior detection. First, the proposed method contains a permutation-equivalent operation to handle sub-trajectories that might have different visiting orders but similar impacts on mobility behaviors. Second, the proposed method utilizes a variational autoencoder architecture to simultaneously perform clustering in both latent and original spaces. Also, in order to handle the bias of a single latent space, our clustering assignment prediction considers multiple learned latent spaces at different epochs. This way, the proposed method produces accurate results and can provide reliability estimates of each trajectory’s cluster assignment. The experiment shows that the proposed method outperformed state-of-the-art methods in mobility behavior detection from trajectories with better accuracy and more interpretability.

Large network logs, recording multivariate time series generated from heterogeneous devices and sensors in a network, can reveal important information about abnormal activities, such as network intrusions and packet losses. Existing machine learning methods for anomaly detection on multiple multivariate time series typically assume that 1) infrequent behaviors beyond some inference threshold are anomalous for unsupervised models or 2) require a large set of labeled normal and abnormal sequences for supervised models. However, in practice, the reported abnormal events might be available but incomplete and sparse (i.e., much fewer than normal cases). This paper presents a novel semi-supervised approach, SNetAD, that takes advantage of the incomplete and imbalanced labels to effectively learn separable feature embeddings of network activities representing normal and abnormal events. Specifically, SNetAD first generates network representations by capturing relationships across time points and between network devices. Then SNetAD encourages the embeddings to form two clusters using contrastive center loss and improves the separability of the learned clusters using labeled and unlabeled samples in a semi-supervised manner. The experiments demonstrate that SNetAD significantly outperforms state-of-the-art approaches for abnormal event prediction on a large real-world network log.

The increasing availability and accessibility of numerous overhead images allows us to estimate and assess the spatial arrangement of groups of geospatial target objects, which can benefit many applications, such as traffic monitoring and agricultural monitoring. Spatial arrangement estimation is the process of identifying the areas which contain the desired objects in overhead images. Traditional supervised object detection approaches can estimate accurate spatial arrangement but require large amounts of bounding box annotations. Recent semi-supervised clustering approaches can reduce manual labeling but still require annotations for all object categories in the image. This paper presents the target-guided generative model (TGGM), under the Variational Auto-encoder (VAE) framework, which uses Gaussian Mixture Models (GMM) to estimate the distributions of both hidden and decoder variables in VAE. Modeling both hidden and decoder variables by GMM reduces the required manual annotations significantly for spatial arrangement estimation. Unlike existing approaches that the training process can only update the GMM as a whole in the optimization iterations (e.g., a "minibatch"), TGGM allows the update of individual GMM components separately in the same optimization iteration. Optimizing GMM components separately allows TGGM to exploit the semantic relationships in spatial data and requires only a few labels to initiate and guide the generative process. Our experiments shows that TGGM achieves results comparable to the state-of-the-art semi-supervised methods and outperformes unsupervised methods by 10% based on the F1scores, while requiring significantly fewer labeled data.

Thousands of scanned historical topographic maps contain valuable information covering long periods of time, such as how the hydrography of a region has changed over time. Efficiently unlocking the information in these maps requires training a geospatial objects recognition system, which needs a large amount of annotated data. Overlapping geo-referenced external vector data with topographic maps according to their coordinates can annotate the desired objects’ locations in the maps automatically. However, directly overlapping the two datasets causes misaligned and false annotations because the publication years and coordinate projection systems of topographic maps are different from the external vector data. We propose a label correction algorithm, which leverages the color information of maps and the prior shape information of the external vector data to reduce misaligned and false annotations. The experiments show that the precision of annotations from the proposed algorithm is 10% higher than the annotations from a state-of-the-art algorithm. Consequently, recognition results using the proposed algorithm’s annotations achieve 9% higher correctness than using the annotations from the state-of-the-art algorithm.

This study proposes a novel approach for location recommendation based on content-based recommendation algorithms incorporated with geographic information. The study also analyzes the impact of various dimension reduction (DR) methods on the recommendation quality using various baseline approaches. The experiment demonstrates that the proposed approach to content-based location recommendations is feasible and valuable, with potentials for further research.

Many scientific prediction problems have spatiotemporal data- and modeling-related challenges in handling complex variations in space and time using only sparse and unevenly distributed observations. This paper presents a novel deep learning architecture, Deep learning predictions for LocATion-dependent Time-sEries data (DeepLATTE), that explicitly incorporates theories of spatial statistics into neural networks to addresses these challenges. In addition to a feature selection module and a spatiotemporal learning module, DeepLATTE contains an autocorrelation-guided semi-supervised learning strategy to enforce both local autocorrelation patterns and global autocorrelation trends of the predictions in the learned spatiotemporal embedding space to be consistent with the observed data, overcoming the limitation of sparse and unevenly distributed observations. During the training process, both supervised and semi-supervised losses guide the updates of the entire network to: 1) prevent overfitting, 2) refine feature selection, 3) learn useful spatiotemporal representations, and 4) improve overall prediction. We conduct a demonstration of DeepLATTE using publicly available data for an important public health topic, air quality prediction, in a well-studied, complex physical environment - Los Angeles. The experiment demonstrates that the proposed approach provides accurate fine-spatial-scale air quality predictions and reveals the critical environmental factors affecting the results.

Historical maps contain detailed geographic information difficult to find elsewhere covering long-periods of time (e.g., 125 years for the historical topographic maps in the US). However, these maps typically exist as scanned images without searchable metadata. Existing approaches making historical maps searchable rely on tedious manual work (including crowd-sourcing) to generate the metadata (e.g., geolocations and keywords). Optical character recognition (OCR) software could alleviate the required manual work, but the recognition results are individual words instead of location phrases (e.g., "Black'' and "Mountain'' vs. "Black Mountain''). This paper presents an end-to-end approach to address the real-world problem of finding and indexing historical map images. This approach automatically processes historical map images to extract their text content and generates a set of metadata that is linked to large external geospatial knowledge bases. The linked metadata in the RDF (Resource Description Framework) format support complex queries for finding and indexing historical maps, such as retrieving all historical maps covering mountain peaks higher than 1,000 meters in California. We have implemented the approach in a system called mapKurator. We have evaluated mapKurator using historical maps from several sources with various map styles, scales, and coverage. Our results show significant improvement over the state-of-the-art methods. The code has been made publicly available as modules of the Kartta Labs project at https://github.com/kartta-labs/Project.

With large amounts of digital map archives becoming available, automatically extracting information from scanned historical maps is needed for many domains that require long-term historical geographic data. Convolutional Neural Networks (CNN) are powerful techniques that can be used for extracting locations of geographic features from scanned maps if sufficient representative training data are available. Existing spatial data can provide the approximate locations of corresponding geographic features in historical maps and thus be useful to annotate training data automatically. However, the feature representations, publication date, production scales, and spatial reference systems of contemporary vector data are typically very different from those of historical maps. Hence, such auxiliary data cannot be directly used for annotation of the precise locations of the features of interest in the scanned historical maps. This research introduces an automatic vector-to-raster alignment algorithm based on reinforcement learning to annotate precise locations of geographic features on scanned maps. This paper models the alignment problem using the reinforcement learning framework, which enables informed, efficient searches for matching features without pre-processing steps, such as extracting specific feature signatures (e.g. road intersections). The experimental results show that our algorithm can be applied to various features (roads, water lines, and railroads) and achieve high accuracy.

Identifying mobility behaviors in rich trajectory data is of great economic and social interest to various applications including urban planning, marketing and intelligence. Existing work on trajectory clustering often relies on similarity measurements that utilize raw spatial and/or temporal information of trajectories. These measures are incapable of identifying similar moving behaviors that exhibit varying spatiotemporal scales of movement. In addition, the expense of labeling massive trajectory data is a barrier to supervised learning models. To address these challenges, we propose an unsupervised neural approach for mobility behavior clustering, called the Deep Embedded TrajEctory ClusTering network (DETECT). DETECT operates in three parts: first it transforms the trajectories by summarizing their critical parts and augmenting them with context derived from their geographical locality (e.g., using POIs from gazetteers). In the second part, it learns a powerful representation of trajectories in the latent space of behaviors, thus enabling a clustering function (such as k-means) to be applied. Finally, a clustering oriented loss is directly built on the embedded features to jointly perform feature refinement and cluster assignment, thus improving separability between mobility behaviors. Exhaustive quantitative and qualitative experiments on two real-world datasets demonstrate the effectiveness of our approach for mobility behavior analyses.

This paper presents an exact solution and a heuristic solution to a UAV-assisted parcel delivery problem, in which UAVs can only be operated in Visual-Line-Of-Sight (VLOS) areas. In our proposed problem, we assume that trucks travel on road networks, and UAVs move in Euclidean spaces and can launch at any locations on roads. We first demonstrate the overview of our exact solution that iterates all permutations of destinations for an optimal delivery route. Given a specific delivery order, an intuitive approach needs to check all possible locations on roads in the VLOS areas and find a globally optimal location for every destination if UAVs are used for delivery. To avoid high computational cost of searching the optimal location at runtime, we propose an advanced index-based alternative, which computes optimal delivery routes in a pre-processing stage. Due to the nature of NP-hard problems, we also propose a heuristic approach that utilizes delivery groups for the proposed problem of practical size. All proposed solutions are evaluated through extensive experiments.

In StarCraft, buildings arrangement near the intersections is one of most the critical strategic decisions in the early stage. The high time complexity of the buildings arrangement makes it difficult for AI bot to make the real-time decision. This paper presents an approach to analyze the intersection in StarCraft II maps. We propose a radarlike algorithm to automatically detect the intersection and use a designed heuristic search algorithm to arrange the building for building the wall. Our method can obtain the optimal solution while meeting the real-time requirement.

Forecasting spatially correlated time series data is challenging because of the linear and non-linear dependencies in the temporal and spatial dimensions. Air quality forecasting is one canonical example of such tasks. Existing work, e.g., auto-regressive integrated moving average (ARIMA) and artificial neural network (ANN), either fails to model the non-linear temporal dependency or cannot effectively consider spatial relationships between multiple spatial time series data. In this paper, we present an approach for forecasting short-term PM2.5 concentrations using a deep learning model, the geo-context based diffusion convolutional recurrent neural network, GC-DCRNN. The model describes the spatial relationship by constructing a graph based on the similarity of the built environment between the locations of air quality sensors. The similarity is computed using the surrounding "important" geographic features regarding their impacts to air quality for each location (e.g., the area size of parks within a 1000-meter buffer, the number of factories within a 500-meter buffer). Also, the model captures the temporal dependency leveraging the sequence to sequence encoder-decoder architecture. We evaluate our model on two real-world air quality datasets and observe consistent improvement of 5%-10% over baseline approaches.

With the widespread installation of location-enabled devices on public transportation, public vehicles are generating massive amounts of trajectory data in real time. However, using these trajectory data for meaningful analysis requires careful considerations in storing, managing, processing, and visualizing the data. Using the location data of the Los Angeles Metro bus system, along with publicly available bus schedule data, we conduct a data processing and analyses study to measure the performance of the public transportation system in Los Angeles utilizing a number of metrics including travel-time reliability, on-time performance, bus bunching, and travel-time estimation. We demonstrate the visualization of the data analysis results through an interactive web-based application. The developed algorithms and system provide powerful tools to detect issues and improve the efficiency of public transportation systems.

Air quality models are important for studying the impact of air pollutant on health conditions at a fine spatiotemporal scale. Existing work typically relies on area-specific, expert-selected attributes of pollution emissions (e,g., transportation) and dispersion (e.g., meteorology) for building the model for each combination of study areas, pollutant types, and spatiotemporal scales. In this paper, we present a data mining approach that utilizes publicly available OpenStreetMap (OSM) data to automatically generate an air quality model for the concentrations of fine particulate matter less than 2.5 μm in aerodynamic diameter at various temporal scales. Our experiment shows that our (domain-) expert-free model could generate accurate PM2.5 concentration predictions, which can be used to improve air quality models that traditionally rely on expert-selected input. Our approach also quantifies the impact on air quality from a variety of geographic features (i.e., how various types of geographic features such as parking lots and commercial buildings affect air quality and from what distance) representing mobile, stationary and area natural and anthropogenic air pollution sources. This approach is particularly important for enabling the construction of context-specific spatiotemporal models of air pollution, allowing investigations of the impact of air pollution exposures on sensitive populations such as children with asthma at scale.

According to the Centers for Disease Control, in the United States there are 6.8 million children living with asthma. Despite the importance of the disease, the available prognostic tools are not sufficient for biomedical researchers to thoroughly investigate the potential risks of the disease at scale. To overcome these challenges we present a big data integration and analysis infrastructure developed by our Data and Software Coordination and Integration Center (DSCIC) of the NIBIB-funded Pediatric Research using Integrated Sensor Monitoring Systems (PRISMS) program. Our goal is to help biomedical researchers to efficiently predict and prevent asthma attacks. The PRISMS-DSCIC is responsible for collecting, integrating, storing, and analyzing real-time environmental, physiological and behavioral data obtained from heterogeneous sensor and traditional data sources. Our architecture is based on the Apache Kafka, Spark and Hadoop frameworks and PostgreSQL DBMS. A main contribution of this work is extending the Spark framework with a mediation layer, based on logical schema mappings and query rewriting, to facilitate data analysis over a consistent harmonized schema. The system provides both batch and stream analytic capabilities over the massive data generated by wearable and fixed sensors.

We present Q2P, a system that discovers query templates from search engines via their query autocompletion services. Q2P is distinct from the existing works in that it does not rely on query logs of search engines that are typically not readily available. Q2P is also unique in that it uses a trie to economically store queries sampled from a search engine and employs a beam-search strategy that focuses the expansion of the trie on its most promising nodes. Furthermore, Q2P leverages the trie-based storage of query sample to discover query templates using only two passes over the trie. Q2P is a key part of our ongoing project Deep2Q on a template-driven data integration on the Deep Web, where the templates learned by Q2P are used to guide the integration process in Deep2Q. Experimental results on four major search engines indicate that (1) Q2P sends only a moderate number of queries (ranging from 597 to 1,135) to the engines, while obtaining a significant number of completions per query (ranging from 4.2 to 8.5 on the average); (2) a significant number of templates (ranging from 8 to 32 when the minimum support for frequent templates is set to 1%) may be discovered from the samples.

Historical spatiotemporal datasets are important for a variety of studies such as cancer and environmental epidemiology, urbanization, and landscape ecology. However, existing data sources typically contain only contemporary datasets. Historical maps hold a great deal of detailed geographic information at various times in the past. Yet, finding relevant maps is difficult and the map content are not machine-readable. I envision a map processing, modeling, linking, and publishing framework that allows querying historical map collections as a unified and structured spatiotemporal source in which individual geographic phenomena (extracted from maps) are modeled with semantic descriptions and linked to other data sources (e.g., DBpedia). This framework will make it possible to efficiently study historical spatiotemporal datasets on a large scale. Realizing such a framework poses significant research challenges in multiple fields in computer science including digital map processing, data integration, and the Semantic Web technologies, and other disciplines such as spatial, earth, social, and health sciences. Tackling these challenges will not only advance research in computer science but also present a unique opportunity for interdisciplinary research.

Map labels provide valuable geographic information by annotating geographic phenomenona with text descriptions. However, many interesting and useful maps are only available as images and hence this information is not readily accessible in a Geographic Information System (GIS). Previous work on text recognition in maps considers maps as a special type of image to be processed using Optical Character Recognition (OCR) techniques and does not pay attention to the typical workflows in a GIS. As a result, to convert map labels into machine-readable text, a user has to switch between OCR and GIS software, transform the detected text locations from the image coordinates (in OCR) to the map coordinates (in GIS), and apply data import/export procedures. This tedious process limits the opportunity to access text information in maps. This paper presents ArcStrabo, an integration of our previous text recognition work and a GIS, which uses a GIS user interface, workflows, and data types to enable efficient training of text recognition algorithms for converting map labels to a table of geographic names. We show that ArcStrabo facilitates map digitization processes, eliminates the need for GIS users to learn additional OCR tools, and does not require manual data export/import procedures between GIS and OCR software.

A significant challenge in handling geographic datasets is that the datasets can come from heterogeneous sources with various data qualities and formats. Before these datasets can be used in a Geographic Information System (GIS) for spatial analysis or to create maps, a typical task is to clean the attribute data and transform the data into a uniform format. However, conventional GIS products focus on manipulating the spatial component of geographic features and only offer basic tools for editing the attribute data (e.g., one row at a time). This limits the capability for handling large datasets in a GIS since manually editing and transforming attribute data between different formats is not practical for thousands of geographic features. In this demo, we present ArcKarma, which is built on our previous work on data transformation, to efficiently clean and transform data attributes in a GIS. ArcKarma generates transformation programs from a few user-provided examples and applies these programs to transform individual attribute columns into the desired formats. We show that ArcKarma produces accurate results and eliminates the need for laborious manual data cleaning and scripting tasks.

Given the increasing popularity and availability of location tracking devices, large quantities of spatiotemporal data are available from many different sources. Quick interactive analysis of such data is important in order to understand the data, identify patterns, and eventually make a marketable product. Since the data do not necessarily follow the relational model and may require flexible processing possibly using advanced machine learning techniques, spatial databases or similar query tools do not make the best means for such analysis. Moreover, the high complexity of geometric operations makes the quick interactive analysis very difficult. In this paper, we present a highly flexible functional query engine that 1) works with multiple schema types, 2) provides fast response times by spatiotemporal indexing and parallelization, 3) helps understand the data using visualizations and 4) is highly extensible to easily add complex functionality. To demonstrate its usefulness, we use our tool to solve a real world problem of crime pattern analysis in Los Angeles County and compare the process with other well known tools.

Text recognition is difficult from documents that contain multi-oriented, curved text lines of various character sizes. This is because layout analysis techniques, which most optical character recognition (OCR) approaches rely on, do not work well on unstructured documents with non-homogeneous text. Previous work on recognizing non-homogeneous text typically handles specific cases, such as horizontal and/or straight text lines and single-sized characters. In this paper, we present a general text recognition technique to handle non-homogeneous text by exploiting dynamic character grouping criteria based on the character sizes and maximum desired string curvature. This technique can be easily integrated with classic OCR approaches to recognize non-homogeneous text. In our experiments, we compared our approach to a commercial OCR product using a variety of raster maps that contain multi-oriented, curved and straight text labels of multi-sized characters. Our evaluation showed that our approach produced accurate text recognition results and outperformed the commercial product at both the word and character level accuracy.

Raster maps contain valuable road information, which is especially important for the areas where road vector data are otherwise not readily accessible. However, converting the road information in raster maps to road vector data usually requires significant user effort to achieve high accuracy. In this demo, we present Strabo, which is a system that extracts road vector data from heterogeneous raster maps. We demonstrate Strabo's fully automatic technique for extracting road vector data from raster maps with good image quality and the semi-automatic technique for handling raster maps with poor image quality. We show that Strabo requires minimal user input for extracting road vector data from raster maps with varying map complexity (i.e., overlapping features in maps) and image quality.

Raster maps are widely available and contain useful geographic features such as labels and road lines. To extract the geographic features, most research work relies on a manual step to first extract the foreground pixels from the maps using the distinctive colors or grayscale intensities of the pixels. This strategy requires user interaction for each map to select a set of thresholds. In this paper, we present a map classification technique that uses an image comparison feature called the luminance-boundary histogram and a nearest-neighbor classifier to identify raster maps with similar grayscale intensity usage. We can then apply previously learned thresholds to separate the foreground pixels from the raster maps that are classified in the same group instead of manually examining each map. We show that the luminance-boundary histogram achieves 95% accuracy in our map classification experiment compared to 13.33%, 86.67%, and 88.33% using three traditional image comparison features. The accurate map classification results make it possible to extract geographic features from previously unseen raster maps.

To exploit the road network in raster maps, the first step is to extract the pixels that constitute the roads and then vectorize the road pixels. Identifying colors that represent roads in raster maps for extracting road pixels is difficult since raster maps often contain numerous colors due to the noise introduced during the processes of image compression and scanning. In this paper, we present an approach that minimizes the required user input for identifying the road colors representing the road network in a raster map. We can then use the identified road colors to extract road pixels from the map. Our approach can be used on scanned and compressed maps that are otherwise difficult to process automatically and tedious to process manually. We tested our approach with 100 maps from a variety of sources, which include 90 scanned maps with various compression levels and 10 computer generated maps. We successfully identified the road colors and extracted the road pixels from all test maps with fewer than four user labels per map on average.

The road network is one of the most important types of information on raster maps. In particular, the set of road intersection templates, which consists of the road intersection positions, the road connectivities, and the road orientations, represents an abstraction of the road network and is more accurate and easier to extract than the extraction of the entire road network. To extract the road intersection templates from raster maps, the thinning operator is commonly used to find the basic structure of the road lines (i.e., to extract the skeletons of the lines). However, the thinning operator produces distorted lines near line intersections, especially at the T-shaped intersections. Therefore, the extracted position of the road intersection and the road orientations are not accurate. In this paper, we utilize our previous work on automatically extracting road intersection positions to identify the road lines that intersect at the intersections and then trace the road orientations and refine the positions of the road intersections. We compare the proposed approach with the usage of the thinning operator and show that our proposed approach extracts more accurate road intersection positions and road orientations than the previous approach.