An‐Bo Li, Hao Chen, Xian‐Li Xie, Guo‐Nian Lü, Matthew Fox
As the research and application of three‐dimensional (3D) Geographic Information Science (GIS) continue to advance, the abstract representation and symbolic modeling of geological entities in three dimensions have become one of the research focuses in the current GIS field. To address the need for the symbolic representation of complex and diverse fault structures, this paper proposes a parametric modeling method for 3D fault symbols. This method includes (1) constructing a 3D stratum model and a fault plane model based on stratum and fault plane parameters, (2) performing a 3D cutting operation based on the fault plane model to generate the fault block model, and (3) translating the strata in two faultblocks according to the parameters of fault motion to generate a fault symbol model. The experimental results show that the proposed method requires only a small number of parameters to efficiently and intuitively construct diverse 3D fault symbol models. This method breaks through the excessive dependence on geological survey data in the process of 3D geological modeling. It is suitable for 3D geological symbol modeling of folds, joints, intrusions, and other geological structures, as well as 3D modeling of typical geological structures with relatively simple spatial morphology. This paper's parametric modeling method has essential research significance and application value in various applications such as digital earth, digital city, and virtual geoscience exploration.
{"title":"Parametric Modeling Method for 3D Symbols of Fault Structures","authors":"An‐Bo Li, Hao Chen, Xian‐Li Xie, Guo‐Nian Lü, Matthew Fox","doi":"10.1111/tgis.13242","DOIUrl":"https://doi.org/10.1111/tgis.13242","url":null,"abstract":"As the research and application of three‐dimensional (3D) Geographic Information Science (GIS) continue to advance, the abstract representation and symbolic modeling of geological entities in three dimensions have become one of the research focuses in the current GIS field. To address the need for the symbolic representation of complex and diverse fault structures, this paper proposes a parametric modeling method for 3D fault symbols. This method includes (1) constructing a 3D stratum model and a fault plane model based on stratum and fault plane parameters, (2) performing a 3D cutting operation based on the fault plane model to generate the fault block model, and (3) translating the strata in two faultblocks according to the parameters of fault motion to generate a fault symbol model. The experimental results show that the proposed method requires only a small number of parameters to efficiently and intuitively construct diverse 3D fault symbol models. This method breaks through the excessive dependence on geological survey data in the process of 3D geological modeling. It is suitable for 3D geological symbol modeling of folds, joints, intrusions, and other geological structures, as well as 3D modeling of typical geological structures with relatively simple spatial morphology. This paper's parametric modeling method has essential research significance and application value in various applications such as digital earth, digital city, and virtual geoscience exploration.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"28 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142196784","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Taking the bike‐sharing travel demand (BSTD) as an example, this study investigates the potential of Nighttime Light (NTL) data to optimize forecasting performance and replace the land use factors. Stepwise regression is trained with the travel demand in each unit as the dependent variable, and land use factors are introduced as the independent variable one by one, which finds the set of independent variables. Five machine learning algorithms driven by ensemble learning and decision trees including the GBDT, Random Forecast, Adaboost, Extratrees, and Catboost, are employed and evaluated to achieve comparative analysis of “before considering‐after considering NTL data”. The methodological verification of Beijing city shows: (1) Adaboost and GBDT are superior to all other algorithms, since they generally have the highest R2, lowest RMSE, and lowest absolute MAPE. (2) All methods by employing NTL data obviously optimize the performance of BSTD forecast with decreased RMSE, decreased MAPE, etc. In particular, GBDT performs the best in reducing MSE, with a percentage of −99.99% in the training set and −86.985% in the test set, which AdaBoost, Extratrees, and Catboost follow. (3) Land use factors no longer make sense in predicting BSTD after employing NTL data, and NTL data has covered the roles of land use factors to ensure accuracy. The conclusions presented here enrich our understanding of the relative roles of land use factors and NTL data in travel demand and boost our optimization in traffic prediction in the future.
{"title":"Investigating the Potential of Nighttime Light Data to Estimate Travel Demand","authors":"Chao Sun, Jian Lu","doi":"10.1111/tgis.13240","DOIUrl":"https://doi.org/10.1111/tgis.13240","url":null,"abstract":"Taking the bike‐sharing travel demand (BSTD) as an example, this study investigates the potential of Nighttime Light (NTL) data to optimize forecasting performance and replace the land use factors. Stepwise regression is trained with the travel demand in each unit as the dependent variable, and land use factors are introduced as the independent variable one by one, which finds the set of independent variables. Five machine learning algorithms driven by ensemble learning and decision trees including the GBDT, Random Forecast, Adaboost, Extratrees, and Catboost, are employed and evaluated to achieve comparative analysis of “before considering‐after considering NTL data”. The methodological verification of Beijing city shows: (1) Adaboost and GBDT are superior to all other algorithms, since they generally have the highest <jats:italic>R</jats:italic><jats:sup>2</jats:sup>, lowest RMSE, and lowest absolute MAPE. (2) All methods by employing NTL data obviously optimize the performance of BSTD forecast with decreased RMSE, decreased MAPE, etc. In particular, GBDT performs the best in reducing MSE, with a percentage of −99.99% in the training set and −86.985% in the test set, which AdaBoost, Extratrees, and Catboost follow. (3) Land use factors no longer make sense in predicting BSTD after employing NTL data, and NTL data has covered the roles of land use factors to ensure accuracy. The conclusions presented here enrich our understanding of the relative roles of land use factors and NTL data in travel demand and boost our optimization in traffic prediction in the future.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"2013 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142196786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Conflict, manifesting as riots and protests, is a common occurrence in urban environments worldwide. Understanding their likely locations is crucial to policymakers, who may (for example) seek to provide overseas travelers with guidance on safe areas, or local policymakers with the ability to pre‐position medical aid or police presences to mediate negative impacts associated with riot events. Past efforts to forecast these events have focused on the use of news and social media, restricting applicability to areas with available data. This study utilizes a ResNet convolutional neural network and high‐resolution satellite imagery to estimate the spatial distribution of riots or protests within urban environments. At a global scale (N = 18,631 conflict events), by training our model to understand relationships between urban form and riot events, we are able to predict the likelihood that a given urban area will experience a riot or protest with accuracy as high as 97%. This research has the potential to improve our ability to forecast and understand the relationship between urban form and conflict events, even in data‐sparse regions.
{"title":"Predicting Protests and Riots in Urban Environments With Satellite Imagery and Deep Learning","authors":"Scott Warnke, Daniel Runfola","doi":"10.1111/tgis.13236","DOIUrl":"https://doi.org/10.1111/tgis.13236","url":null,"abstract":"Conflict, manifesting as riots and protests, is a common occurrence in urban environments worldwide. Understanding their likely locations is crucial to policymakers, who may (for example) seek to provide overseas travelers with guidance on safe areas, or local policymakers with the ability to pre‐position medical aid or police presences to mediate negative impacts associated with riot events. Past efforts to forecast these events have focused on the use of news and social media, restricting applicability to areas with available data. This study utilizes a ResNet convolutional neural network and high‐resolution satellite imagery to estimate the spatial distribution of riots or protests within urban environments. At a global scale (<jats:italic>N</jats:italic> = 18,631 conflict events), by training our model to understand relationships between urban form and riot events, we are able to predict the likelihood that a given urban area will experience a riot or protest with accuracy as high as 97%. This research has the potential to improve our ability to forecast and understand the relationship between urban form and conflict events, even in data‐sparse regions.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"9 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142196785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the widespread worldwide adoption of location‐based service technologies, accurate and reliable driving trajectories have become crucial. However, because of the inherent deficiencies of sensor devices, accurate road matching results may not always be obtained directly from trajectory data, which poses a challenge for many location and trajectory based services. Existing map‐matching techniques mainly focus on high‐sampling‐rate trajectory data while paying relatively less attention to low‐frequency trajectory data. Low‐sampling‐rate trajectory data have greater matching difficulties than high‐sampling‐rate data owing to the limited available information. Moreover, in the case of signal loss or interference, the accuracy of map‐matching algorithms can decrease significantly for low‐sampling‐rate data. To achieve accurate map‐matching results for low‐sampling‐rate trajectory data, this study proposes a map‐matching algorithm based on probability interpolation. First, the trajectory data are denoised to eliminate redundant trajectory points. Second, the concept of the probability truth value is introduced to handle the relationship between the interpolated virtual points and actual sampled trajectory points accurately. A higher probability truth value indicates a higher confidence level of the interpolation. Third, the denoised trajectory data are interpolated and a probability truth value is assigned based on the interpolation accuracy. Finally, a comprehensive probability composed of the probability truth value, emission probability, and transition probability is used to determine the correctly matched road segments. Experimental results on real trajectory datasets demonstrated that the proposed algorithm outperformed several advanced algorithms in terms of accuracy and performance.
{"title":"Low‐Frequency Trajectory Map‐Matching Method Based on Probability Interpolation","authors":"Wenkai Wang, Qingying Yu, Ruijia Duan, Qi Jin, Xiang Deng, Chuanming Chen","doi":"10.1111/tgis.13234","DOIUrl":"https://doi.org/10.1111/tgis.13234","url":null,"abstract":"With the widespread worldwide adoption of location‐based service technologies, accurate and reliable driving trajectories have become crucial. However, because of the inherent deficiencies of sensor devices, accurate road matching results may not always be obtained directly from trajectory data, which poses a challenge for many location and trajectory based services. Existing map‐matching techniques mainly focus on high‐sampling‐rate trajectory data while paying relatively less attention to low‐frequency trajectory data. Low‐sampling‐rate trajectory data have greater matching difficulties than high‐sampling‐rate data owing to the limited available information. Moreover, in the case of signal loss or interference, the accuracy of map‐matching algorithms can decrease significantly for low‐sampling‐rate data. To achieve accurate map‐matching results for low‐sampling‐rate trajectory data, this study proposes a map‐matching algorithm based on probability interpolation. First, the trajectory data are denoised to eliminate redundant trajectory points. Second, the concept of the probability truth value is introduced to handle the relationship between the interpolated virtual points and actual sampled trajectory points accurately. A higher probability truth value indicates a higher confidence level of the interpolation. Third, the denoised trajectory data are interpolated and a probability truth value is assigned based on the interpolation accuracy. Finally, a comprehensive probability composed of the probability truth value, emission probability, and transition probability is used to determine the correctly matched road segments. Experimental results on real trajectory datasets demonstrated that the proposed algorithm outperformed several advanced algorithms in terms of accuracy and performance.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"192 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142225059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The COVID‐19 pandemic, which originated in China at the end of 2019, escalated into a global crisis by March 2020. To mitigate its spread, governments worldwide implemented strict lockdown measures. While these lockdowns had adverse social, economic, and health impacts, they also led to significant environmental improvements in many regions. India's urban environment also significantly improved during lockdown. This study investigates the changes in Land Surface Temperature (LST) across eight major Indian cities, each representing diverse climatic and physiographic zones: Delhi, Dehradun, Lucknow, Kolkata, Bhopal, Bhubaneshwar, Mumbai, and Hyderabad. It aims to enhance the understanding of how sudden reductions in anthropogenic activities influence urban temperatures. The LST was computed for the lockdown period of April to May 2020 and was compared with the pre‐lockdown years of 2018 and 2019 and the post‐lockdown year of 2021, utilizing Landsat thermal data processed through the mono‐window algorithm. The results exhibit significant reductions in LST during the lockdown period. Cities like Delhi, Dehradun, and Lucknow experienced a reduction of 6°C, 5°C, and 4°C, respectively, in LST from pre‐lockdown to lockdown periods. In contrast, cities like Bhopal, Bhubaneswar, Mumbai, and Hyderabad experienced a reduction of around 2°C–3°C. However, the city of Kolkata showed an increase of 3°C from 2019 to 2020. These results highlight the substantial influence of human activities on urban thermal environments and underline the potential benefits of reducing anthropogenic impacts to improve urban thermal well‐being.
{"title":"Land Surface Temperature Dynamics during COVID‐19 Lockdown in Diverse Climatic and Physiographic Zones—A Study of Indian Mega Cities","authors":"Ashish Mishra, Dhyan S. Arya","doi":"10.1111/tgis.13237","DOIUrl":"https://doi.org/10.1111/tgis.13237","url":null,"abstract":"The COVID‐19 pandemic, which originated in China at the end of 2019, escalated into a global crisis by March 2020. To mitigate its spread, governments worldwide implemented strict lockdown measures. While these lockdowns had adverse social, economic, and health impacts, they also led to significant environmental improvements in many regions. India's urban environment also significantly improved during lockdown. This study investigates the changes in Land Surface Temperature (LST) across eight major Indian cities, each representing diverse climatic and physiographic zones: Delhi, Dehradun, Lucknow, Kolkata, Bhopal, Bhubaneshwar, Mumbai, and Hyderabad. It aims to enhance the understanding of how sudden reductions in anthropogenic activities influence urban temperatures. The LST was computed for the lockdown period of April to May 2020 and was compared with the pre‐lockdown years of 2018 and 2019 and the post‐lockdown year of 2021, utilizing Landsat thermal data processed through the mono‐window algorithm. The results exhibit significant reductions in LST during the lockdown period. Cities like Delhi, Dehradun, and Lucknow experienced a reduction of 6°C, 5°C, and 4°C, respectively, in LST from pre‐lockdown to lockdown periods. In contrast, cities like Bhopal, Bhubaneswar, Mumbai, and Hyderabad experienced a reduction of around 2°C–3°C. However, the city of Kolkata showed an increase of 3°C from 2019 to 2020. These results highlight the substantial influence of human activities on urban thermal environments and underline the potential benefits of reducing anthropogenic impacts to improve urban thermal well‐being.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"4 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142196787","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Unlike the simplification of individual lines, the generalization of contour clusters oriented to geographical features is a structured generalization behavior that extracts knowledge of geographical features form the perspective of spatial cognition. In decision level, terrain generalization is essentially a similarity transformation between the geomorphic structures corresponding to multi‐scale contour cluster. However, the multi‐scale structural similarity relations are not directly connected with the application of contour generalization. Therefore, this paper presents an automated method for terrain contour structured generalization considering multi‐scale structural similarity. Firstly, a drainage tree structure is constructed from contour lines to establish associations between valley branches and contour bends. Then, the quantitative relationships between multi‐scale structural similarity and map scale changes are explored using an indirect quantitative expression method. Finally, the contour structural generalization is fully automated through iterative optimization principle based on the multi‐scale structural similarity relations. The experiment results demonstrate the rationality and feasibility of fully automating the contour generalization process based on multi‐scale geomorphic structural similarity relations. And the proposed method not only overcomes the challenge of determining “how much to select” in map generalization, but also is valuable for enriching the content of spatial similarity relations and map generalization, thereby providing a theoretical method system and support for the construction of national basic vector databases.
{"title":"Automated Generalization of Contour Cluster Considering Multi‐Scale Structural Similarity Relations","authors":"Rong Wang, Haowen Yan, Juanli Jin, Xiaorong Gao","doi":"10.1111/tgis.13232","DOIUrl":"https://doi.org/10.1111/tgis.13232","url":null,"abstract":"Unlike the simplification of individual lines, the generalization of contour clusters oriented to geographical features is a structured generalization behavior that extracts knowledge of geographical features form the perspective of spatial cognition. In decision level, terrain generalization is essentially a similarity transformation between the geomorphic structures corresponding to multi‐scale contour cluster. However, the multi‐scale structural similarity relations are not directly connected with the application of contour generalization. Therefore, this paper presents an automated method for terrain contour structured generalization considering multi‐scale structural similarity. Firstly, a drainage tree structure is constructed from contour lines to establish associations between valley branches and contour bends. Then, the quantitative relationships between multi‐scale structural similarity and map scale changes are explored using an indirect quantitative expression method. Finally, the contour structural generalization is fully automated through iterative optimization principle based on the multi‐scale structural similarity relations. The experiment results demonstrate the rationality and feasibility of fully automating the contour generalization process based on multi‐scale geomorphic structural similarity relations. And the proposed method not only overcomes the challenge of determining “how much to select” in map generalization, but also is valuable for enriching the content of spatial similarity relations and map generalization, thereby providing a theoretical method system and support for the construction of national basic vector databases.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"185 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142227770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chen‐Chieh Feng, Wei Chien Benny Chin, Siyao Gao, Vincent Chua, Elaine Lynn‐Ee Ho
Understanding the interactions between older adults and their living spaces is an important research topic within the human dynamics of geographical scholarship because of their implications on the quality of aging. The prevailing theory of aging tends to stress aging in place which associates aging well with the home and community, thus inadvertently privileging residential neighborhoods as the focal point where older adults spend most of their time. Recent studies have, however, suggested that older adults also age in (social) networks that extend beyond their neighborhoods' physical locations. The need to incorporate perspectives on aging in networks has challenged the prioritization and the enforcement of absolute (or physical) space in GIS software. This study investigates the potential for integrating four top‐level spaces (absolute, relative, relational, and mental) to capture different dimensions of older adults' everyday lives. By leveraging on a dataset of older adults in Singapore, this study demonstrates how the four spaces combined facilitates understanding of aging, specifically their activity spaces, that would otherwise be achieved in a fragmented manner.
{"title":"Illustrating a Splatial Framework to Aging: Absolute, Relative, Relational, and Mental Space in Singapore","authors":"Chen‐Chieh Feng, Wei Chien Benny Chin, Siyao Gao, Vincent Chua, Elaine Lynn‐Ee Ho","doi":"10.1111/tgis.13235","DOIUrl":"https://doi.org/10.1111/tgis.13235","url":null,"abstract":"Understanding the interactions between older adults and their living spaces is an important research topic within the human dynamics of geographical scholarship because of their implications on the quality of aging. The prevailing theory of aging tends to stress aging in place which associates aging well with the home and community, thus inadvertently privileging residential neighborhoods as the focal point where older adults spend most of their time. Recent studies have, however, suggested that older adults also age in (social) networks that extend beyond their neighborhoods' physical locations. The need to incorporate perspectives on aging in networks has challenged the prioritization and the enforcement of absolute (or physical) space in GIS software. This study investigates the potential for integrating four top‐level spaces (absolute, relative, relational, and mental) to capture different dimensions of older adults' everyday lives. By leveraging on a dataset of older adults in Singapore, this study demonstrates how the four spaces combined facilitates understanding of aging, specifically their activity spaces, that would otherwise be achieved in a fragmented manner.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"16 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142196790","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Melissa Meyer, Joe Weber, Selima Sultana, Wanyun Shao
Although comprised of America's most iconic and varied ecosystems and landmarks, the boundaries of national parks have received little attention. This study uses boundary data from the National Park Service and other sources with compactness measures calculated using Geographic Information Systems. Using two common measures of compactness, the Polsby‐Popper and Reock methods, this study aims to answer several research questions: how are the type, location, and age of park units related to compactness, and how are changes in park boundaries related to changes in compactness? Compactness was found to vary between the types of national park units, as well as based on the location of parks within the country. Individual parks have become less compact over time. Due to the significance of national parks to ecosystem conservation, results of this study have crucial implications, providing some direction for future studies and federal regulation as well as overall conservation effort.
{"title":"Geographic Analysis of the Compactness of National Park Unit Boundaries","authors":"Melissa Meyer, Joe Weber, Selima Sultana, Wanyun Shao","doi":"10.1111/tgis.13238","DOIUrl":"https://doi.org/10.1111/tgis.13238","url":null,"abstract":"Although comprised of America's most iconic and varied ecosystems and landmarks, the boundaries of national parks have received little attention. This study uses boundary data from the National Park Service and other sources with compactness measures calculated using Geographic Information Systems. Using two common measures of compactness, the Polsby‐Popper and Reock methods, this study aims to answer several research questions: how are the type, location, and age of park units related to compactness, and how are changes in park boundaries related to changes in compactness? Compactness was found to vary between the types of national park units, as well as based on the location of parks within the country. Individual parks have become less compact over time. Due to the significance of national parks to ecosystem conservation, results of this study have crucial implications, providing some direction for future studies and federal regulation as well as overall conservation effort.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"1 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142196788","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Generative AI including large language models (LLMs) has recently gained significant interest in the geoscience community through its versatile task‐solving capabilities including programming, arithmetic reasoning, generation of sample data, time‐series forecasting, toponym recognition, or image classification. Existing performance assessments of LLMs for spatial tasks have primarily focused on ChatGPT, whereas other chatbots received less attention. To narrow this research gap, this study conducts a zero‐shot correctness evaluation for a set of 76 spatial tasks across seven task categories assigned to four prominent chatbots, that is, ChatGPT‐4, Gemini, Claude‐3, and Copilot. The chatbots generally performed well on tasks related to spatial literacy, GIS theory, and interpretation of programming code and functions, but revealed weaknesses in mapping, code writing, and spatial reasoning. Furthermore, there was a significant difference in the correctness of results between the four chatbots. Responses from repeated tasks assigned to each chatbot showed a high level of consistency in responses with matching rates of over 80% for most task categories in the four chatbots.
{"title":"Correctness Comparison of ChatGPT‐4, Gemini, Claude‐3, and Copilot for Spatial Tasks","authors":"Hartwig H. Hochmair, Levente Juhász, Takoda Kemp","doi":"10.1111/tgis.13233","DOIUrl":"https://doi.org/10.1111/tgis.13233","url":null,"abstract":"Generative AI including large language models (LLMs) has recently gained significant interest in the geoscience community through its versatile task‐solving capabilities including programming, arithmetic reasoning, generation of sample data, time‐series forecasting, toponym recognition, or image classification. Existing performance assessments of LLMs for spatial tasks have primarily focused on ChatGPT, whereas other chatbots received less attention. To narrow this research gap, this study conducts a zero‐shot correctness evaluation for a set of 76 spatial tasks across seven task categories assigned to four prominent chatbots, that is, ChatGPT‐4, Gemini, Claude‐3, and Copilot. The chatbots generally performed well on tasks related to spatial literacy, GIS theory, and interpretation of programming code and functions, but revealed weaknesses in mapping, code writing, and spatial reasoning. Furthermore, there was a significant difference in the correctness of results between the four chatbots. Responses from repeated tasks assigned to each chatbot showed a high level of consistency in responses with matching rates of over 80% for most task categories in the four chatbots.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"14 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142196789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhe Zhang, Jerad King, Shaowen Wang, Diana Sinton, John Wilson, Eric Shook
Maintaining educational resources and training materials as timely, current, and aligned with the needs of students, practitioners, and other users of geospatial technologies is a persistent challenge. This is particularly problematic within CyberGIS, a subfield of Geographic Information Science and Technology (GIS&T) that involves high‐performance computing and advanced cyberinfrastructure to address computation‐ and data‐intensive problems. In this study, we analyzed and compared content from two open educational resources: (1) a popular online web resource that regularly covers CyberGIS‐related topics (GIS Stack Exchange) and (2) existing and proposed content in the GIS&T Body of Knowledge. While current curricula may build a student's conceptual understanding of CyberGIS, there is a noticeable lack of resources for practical implementation of CyberGIS tools. The results highlight discrepancies between the attention and frequency of CyberGIS topics according to a popular online help resource and the CyberGIS academic community.
{"title":"Moving CyberGIS education forward: Knowing what matters and how it is decided","authors":"Zhe Zhang, Jerad King, Shaowen Wang, Diana Sinton, John Wilson, Eric Shook","doi":"10.1111/tgis.13225","DOIUrl":"https://doi.org/10.1111/tgis.13225","url":null,"abstract":"Maintaining educational resources and training materials as timely, current, and aligned with the needs of students, practitioners, and other users of geospatial technologies is a persistent challenge. This is particularly problematic within CyberGIS, a subfield of Geographic Information Science and Technology (GIS&T) that involves high‐performance computing and advanced cyberinfrastructure to address computation‐ and data‐intensive problems. In this study, we analyzed and compared content from two open educational resources: (1) a popular online web resource that regularly covers CyberGIS‐related topics (GIS Stack Exchange) and (2) existing and proposed content in the GIS&T Body of Knowledge. While current curricula may build a student's conceptual understanding of CyberGIS, there is a noticeable lack of resources for practical implementation of CyberGIS tools. The results highlight discrepancies between the attention and frequency of CyberGIS topics according to a popular online help resource and the CyberGIS academic community.","PeriodicalId":47842,"journal":{"name":"Transactions in GIS","volume":"59 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141940406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}