During a natural disaster, people post text messages on various platforms, such as social media and short message service (SMS) platforms, to share urgent information and seek help. Many text messages contain location descriptions about victims and accidents. Accurately extracting these location descriptions can help disaster responders reach victims more quickly and even save lives. These location descriptions, however, are often more complex than simple place names (e.g. city names), and cannot be extracted using typical named entity recognition approaches. While new machine learning models could be trained, they require labeled training data that are time-consuming to create without an effective data annotation tool. To fill this gap, we develop GALLOC, a GeoAnnotator for Labeling LOCation descriptions from disaster-related text messages. GALLOC is an open-source and Web-based tool that provides a variety of functions for supporting location description annotation, such as artificial intelligence powered pre-annotation and automatic spatial footprint identification. It also supports multilingual data annotation, and can be used by a group of users to collaboratively create a dataset. We present the design considerations and functions of GALLOC and evaluate it via a comparison with previous tools and an experiment to annotate a small set of disaster-related messages.

Social media messages posted by people during natural disasters often contain important location descriptions, such as the locations of victims. Recent research has shown that many of these location descriptions go beyond simple place names, such as city names and street names, and are difficult to extract using typical named entity recognition (NER) tools. While advanced machine learning models could be trained, they require large labeled training datasets that can be time-consuming and labor-intensive to create. In this work, we propose a method that fuses geo-knowledge of location descriptions and a Generative Pre-trained Transformer (GPT) model, such as ChatGPT and GPT-4. The result is a geo-knowledge-guided GPT model that can accurately extract location descriptions from disaster-related social media messages. Also, only 22 training examples encoding geo-knowledge are used in our method. We conduct experiments to compare this method with nine alternative approaches on a dataset of tweets from Hurricane Harvey. Our method demonstrates an over 40% improvement over typically used NER approaches. The experiment results also show that geo-knowledge is indispensable for guiding the behavior of GPT models. The extracted location descriptions can help disaster responders reach victims more quickly and may even save lives.

Large pre-trained models, also known as foundation models (FMs), are trained in a task-agnostic manner on large-scale data and can be adapted to a wide range of downstream tasks by fine tuning, few-shot, or even zero-shot learning. Despite their successes in language and vision tasks, we have yet to see an attempt to develop foundation models for geospatial artificial intelligence (GeoAI). In this work, we explore the promises and challenges for developing multimodal foundation models for GeoAI. We first show the advantages of this idea by testing the performance of existing Large pre-trained Language Models (LLMs) (e.g. GPT-2 and GPT-3) on two geospatial semantics tasks. Results indicate that these task-agnostic LLMs can outperform task-specific fully-supervised models on both tasks with 2--9% improvement in a few-shot learning setting. However, we also show the limitations of these existing foundation models given the multimodality nature of GeoAI, especially when dealing with geometries in conjunction with other modalities. So we discuss the possibility of a multimodal foundation model which can reason over various types of geospatial data through geospatial alignments. We conclude this paper by discussing the unique risks and challenges to develop such model for GeoAI.

Maps in the form of digital images are widely available in geoportals, Web pages, and other data sources. The metadata of map images, such as spatial extents and place names, are critical for their indexing and searching. However, many map images have either mismatched metadata or no metadata at all. Recent developments in deep learning offer new possibilities for enriching the metadata of map images via image-based information extraction. One major challenge of using deep learning models is that they often require large amounts of training data that have to be manually labeled. To address this challenge, this paper presents a deep learning approach with GIS-based data augmentation that can automatically generate labeled training map images from shapefiles using GIS operations. We utilize such an approach to enrich the metadata of map images by adding spatial extents and place names extracted from map images. We evaluate this GIS-based data augmentation approach by using it to train multiple deep learning models and testing them on two different datasets: a Web Map Service image dataset at the continental scale and an online map image dataset at the state scale. We then discuss the advantages and limitations of the proposed approach.

A common need for artificial intelligence models in the broader geoscience is to encode various types of spatial data, such as points, polylines, polygons, graphs, or rasters, in a hidden embedding space so that they can be readily incorporated into deep learning models. One fundamental step is to encode a single point location into an embedding space, such that this embedding is learning-friendly for downstream machine learning models. We call this process location encoding. However, there lacks a systematic review on location encoding, its potential applications, and key challenges that need to be addressed. This paper aims to fill this gap. We first provide a formal definition of location encoding, and discuss the necessity of it for GeoAI research. Next, we provide a comprehensive survey about the current landscape of location encoding research. We classify location encoding models into different categories based on their inputs and encoding methods, and compare them based on whether they are parametric, multi-scale, distance preserving, and direction aware. We demonstrate that existing location encoders can be unified under one formulation framework. We also discuss the application of location encoding. Finally, we point out several challenges that need to be solved in the future.

Kai Sunabc , Yingjie Huc , Jia Songad & Yunqiang Zhuad* a State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, Chinab University of Chinese Academy of Sciences, Beijing, Chinac GeoAI Lab, Department of Geography, University at Buffalo, Buffalo, NY, USAd Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application, Nanjing, ChinaKai Sun is a PhD student at the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences. He is also a visiting PhD student in the Department of Geography, University at Buffalo. His research interest is in geospatial semantics. He extracts information from data and builds knowledge base to make the information easier to be accessed. He also develops methods to align geospatial information to deal with the issues of data duplication and inconsistency among different geospatial knowledge bases.Dr. Yingjie Hu is an Assistant Professor in the Department of Geography at the University at Buffalo (UB) and the National Center for Geographic Information and Analysis (NCGIA). His major research area is in geographic information science (GIScience), and more specifically in geospatial artificial intelligence (GeoAI), spatial data mining, and geographic information retrieval. He develops and applies spatial analysis, data mining, machine learning, and deep learning methods to address various geospatial problems in disaster response, public health, urban planning, and digital humanities.Dr. Jia Song is an Associate Professor at the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences. His main research interests include geospatial computing and big data processing.Dr. Yunqiang Zhu is a Professor at the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences. He is interested in studying the spatial information technology research and application, scientific data sharing, and e-Geoscience.CONTACT Yunqiang Zhu [email protected] maps contain rich geographic information about the past of a region. They are sometimes the only source of information before the availability of digital maps. Despite their valuable content, it is often challenging to access and use the information in historical maps, due to their forms of paper-based maps or scanned images. It is even more time-consuming and labor-intensive to conduct an analysis that requires a synthesis of the information from multiple historical maps. To facilitate the use of the geographic information contained in historical maps, one way is to build a geographic knowledge graph (GKG) from them. This paper proposes a general workflow for completing one important step of building such a GKG, namely aligning the same geographic entities from different maps. We present this workflow and the related methods for implementation, and systematically evaluate their performances using two different datasets of historical maps. The evaluation results show that machine learning and deep learning models for matching place names are sensitive to the thresholds learned from the training data, and a combination of measures based on string similarity, spatial distance, and approximate topological relation achieves the best performance with an average F-score of 0.89.

Recent progress in Artificial Intelligence (AI) techniques, the large-scale availability of high-quality data, as well as advances in both hardware and software to efficiently process these data, a...

Local place names are frequently used by residents living in a geographic region. Such place names may not be recorded in existing gazetteers, due to their vernacular nature, relative insignificance to a gazetteer covering a large area (e.g. the entire world), recent establishment (e.g. the name of a newly-opened shopping center) or other reasons. While not always recorded, local place names play important roles in many applications, from supporting public participation in urban planning to locating victims in disaster response. In this paper, we propose a computational framework for harvesting local place names from geotagged housing advertisements. We make use of those advertisements posted on local-oriented websites, such as Craigslist, where local place names are often mentioned. The proposed framework consists of two stages: natural language processing (NLP) and geospatial clustering. The NLP stage examines the textual content of housing advertisements and extracts place name candidates. The geospatial stage focuses on the coordinates associated with the extracted place name candidates and performs multiscale geospatial clustering to filter out the non-place names. We evaluate our framework by comparing its performance with those of six baselines. We also compare our result with four existing gazetteers to demonstrate the not-yet-recorded local place names discovered by our framework.

Cognitive regions and places are notoriously difficult to represent in geographic information science and systems. The exact delineation of cognitive regions is challenging insofar as borders are vague, membership within the regions varies non-monotonically, and raters cannot be assumed to assess membership consistently and homogeneously. In a study published in this journal in 2014, researchers devised a novel grid-based task in which participants rated the membership of individual cells in a given region and contrasted this approach to a standard boundary-drawing task. Specifically, the authors assessed the vague cognitive regions of Northern California and Southern California. The boundary between these cognitive regions was found to have variable width, and region membership peaked not at the most northern or southern cells but at substantially less extreme latitudes. The authors thus concluded that region membership is about attitude, not just latitude. In the present work, we reproduce this study by approaching it from a computational fourth-paradigm perspective, i.e., by the synthesis of high volumes of heterogeneous data from various sources. We compare the regions which we identify to those from the human-participants study of 2014, identifying differences and commonalities. Our results show a significant positive correlation to those in the original study. Beyond the extracted regions themselves, we compare and contrast the empirical and analytical approaches of these two methods, one a conventional human-participants study and the other an application of increasingly popular data-synthesis-driven research methods in GIScience.

In this work we introduce an anisotropic density-based clustering algorithm. It outperforms DBSCAN and OPTICS for the detection of anisotropic spatial point patterns and performs equally well in cases that do not explicitly benefit from an anisotropic perspective. ADCN has the same time complexity as DBSCAN and OPTICS, namely O(n log n) when using a spatial index, O(n2) otherwise.

Recent years have witnessed a large increase in the amount of information available from the Web and many other sources. Such an information deluge presents a challenge for individuals who have to identify useful information items to complete particular tasks in hand. Information value theory (IVT) from economics and artificial intelligence has provided some guidance on this issue. However, existing IVT studies often focus on monetary values, while ignoring the spatiotemporal properties which can play important roles in everyday tasks. In this paper, we propose a theoretical framework for task-oriented information value measurement. This framework integrates IVT with the space-time prism from time geography and measures the value of information based on its impact on an individual’s space-time prisms and its capability of improving task planning. We develop and formalize this framework by extending the utility function from space-time accessibility studies and elaborate it using a simplified example from time geography. We conduct a simulation on a real-world transportation network using the proposed framework. Our research could be applied to improving information display on small-screen mobile devices (e.g., smartwatches) by assigning priorities to different information items.

The research field of scientometrics is concerned with measuring and analyzing science. In practice, this is often done by restricting the impact of publications, journals, and researchers to a mere frequency. However, scientific activities (co-publication, citation, labor mobility) display clear spatiotemporal patterns, and such patterns have rarely been considered in traditional scientometrics. In this work we focus on the study of citations and present a spatiotemporal scientometrics framework to measure the citation impact of research output by taking physical space, place, and time into account. Specifically, we use the statistics of categorical places (institutions, cities, and countries), spatiotemporal kernel density estimations, cartograms, distance distribution curves, and point-pattern analysis to identify spatiotemporal citation patterns. Moreover, we propose a series of s-indices, such as S_institution-index, S_city-index, and S_country-index to evaluate a scientist's impact as a complement to non-spatial citation indicators, e.g., h-index and g-index. In addition, we have developed an interactive web application which allows users to visually explore research topics, authors, publications, as well as the spread of citations through space and time. Our work offers insights on the role of location in scientific knowledge diffusion.

The Semantic Web journal by IOS Press follows a unique open and transparent process during which each submitted manuscript is available online together with the full history of its successive decision statuses, assigned editors, solicited and voluntary reviewers, their full text reviews, and in many cases also the authors’ response letters. Combined with a highly-customized, Drupal-based journal management system, this provides the journal with semantically rich manuscript time lines and networked data about authors, reviewers, and editors. These data are now exposed using a SPARQL endpoint, an extended Bibo ontology, and a modular Linked Data portal that provides interactive scientometrics based on established and new analysis methods. The portal can be customized for other journals as well. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Semantic desktops are a novel approach to improve user interfaces by recording, semantically annotating, and learning from the user's activities to create a personalized user experience and improve search. Such activities, however, are restricted to the information universe, i.e., they only cover events on the local desktop. A next step towards smart mobile devices is the integration of those desktop events with the user's activities in the physical world. Establishing such mappings enables the device to draw conclusions from the recorded desktop events to those that the user is likely performing in the physical world. A Personal Information Management (PIM) system can then better assist the user in task planning and routing. In this work, we propose activity ontologies as blueprints to model the user's activities in the physical world, and use these ontologies to link the Semantic Desktop and the information available on the Web of Linked Data. We discuss the principles of designing the activity ontologies and how to employ them to associate local files and applications with complementary information from the Web. We design a specific activity ontology for a conference use case and present a user interface that extends the Zeitgeist Semantic Desktop to evaluate our approach.