Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)最新文献
Paul Rayson, Alexander Reinhold, J. Butler, Christopher Donaldson, I. Gregory, Joanna E. Taylor
This paper describes the development of an annotated corpus which forms a challenging testbed for geographical text analysis methods. This dataset, the Corpus of Lake District Writing (CLDW), consists of 80 manually digitised and annotated texts (comprising over 1.5 million word tokens). These texts were originally composed between 1622 and 1900, and they represent a range of different genres and authors. Collectively, the texts in the CLDW constitute an indicative sample of writing about the English Lake District during the early seventeenth century and the early twentieth century. The corpus is annotated more deeply than is currently possible with vanilla Named Entity Recognition, Disambiguation and geoparsing. This is especially true of the geographical information the corpus contains, since we have undertaken not only to link different historical and spelling variants of place-names, but also to identify and to differentiate geographical features such as waterfalls, woodlands, farms or inns. In addition, we illustrate the potential of the corpus as a gold standard by evaluating the results of three different NLP libraries and geoparsers on its contents. In the evaluation, the standard NER processing of the text by the different NLP libraries produces many false positive and false negative results, showing the strength of the gold standard.
{"title":"A deeply annotated testbed for geographical text analysis: The Corpus of Lake District Writing","authors":"Paul Rayson, Alexander Reinhold, J. Butler, Christopher Donaldson, I. Gregory, Joanna E. Taylor","doi":"10.1145/3149858.3149865","DOIUrl":"https://doi.org/10.1145/3149858.3149865","url":null,"abstract":"This paper describes the development of an annotated corpus which forms a challenging testbed for geographical text analysis methods. This dataset, the Corpus of Lake District Writing (CLDW), consists of 80 manually digitised and annotated texts (comprising over 1.5 million word tokens). These texts were originally composed between 1622 and 1900, and they represent a range of different genres and authors. Collectively, the texts in the CLDW constitute an indicative sample of writing about the English Lake District during the early seventeenth century and the early twentieth century. The corpus is annotated more deeply than is currently possible with vanilla Named Entity Recognition, Disambiguation and geoparsing. This is especially true of the geographical information the corpus contains, since we have undertaken not only to link different historical and spelling variants of place-names, but also to identify and to differentiate geographical features such as waterfalls, woodlands, farms or inns. In addition, we illustrate the potential of the corpus as a gold standard by evaluating the results of three different NLP libraries and geoparsers on its contents. In the evaluation, the standard NER processing of the text by the different NLP libraries produces many false positive and false negative results, showing the strength of the gold standard.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"12 1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82833658","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Our project involves building a platform able to retrieve, map and analyze the occurrences of place names in fictional novels published between 1800 and 1914 and whose action occurs wholly or partly in Paris. We describe a proof of concept using queries made via the TXM textual analysis platform for the extraction of street names. Then, we propose a fully automatic process using the named entity recognition (NER) components of the PERDIDO platform. This paper describes some encouraging initial results obtained by combining NLP approaches (NER methods) with textometric tools for the automated geoparsing of street names.
{"title":"Automated Geoparsing of Paris Street Names in 19th Century Novels","authors":"Ludovic Moncla, M. Gaio, T. Joliveau, Y. L. Lay","doi":"10.1145/3149858.3149859","DOIUrl":"https://doi.org/10.1145/3149858.3149859","url":null,"abstract":"Our project involves building a platform able to retrieve, map and analyze the occurrences of place names in fictional novels published between 1800 and 1914 and whose action occurs wholly or partly in Paris. We describe a proof of concept using queries made via the TXM textual analysis platform for the extraction of street names. Then, we propose a fully automatic process using the named entity recognition (NER) components of the PERDIDO platform. This paper describes some encouraging initial results obtained by combining NLP approaches (NER methods) with textometric tools for the automated geoparsing of street names.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"138 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73100874","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
C. Chagnaud, Philippe Garat, Paule-Annick Davoine, Elisabetta Carpitelli, Axel Vincent
For decades, geolinguists have been using cartographic materials to display their data and understand the spatial distribution of local dialects. They have used many forms of maps such as maps with labels, symbols or colors. The most widely used are maps with isoglosses which are hand-drawn boundaries defining areas where people share the same language feature. Our source data is a mesh of georeferenced survey points representing a region of interest and each point of the mesh has a phonetic form attached. The issue is to transform this survey point mesh into homogeneous areas sharing similar categorical values in order to identify spatial patterns. This paper describes interpolation methods implemented to produce isogloss maps, namely turning a sample of points into areas boundaries. Implementing such methods requires performing spatial interpolations on a qualitative data set. We also describe a new cartographic tool, ShinyDialect, made for geolinguists to automate the construction process of maps with isoglosses in order to use it as support for spatial analysis. First, we will discuss geolinguistic data and the limits of the existing methods to compute linguistic areas. Next, we will describe the spatial interpolation methods we have implemented. Lastly, we will present features of the tool ShinyDialect to help geolinguists to build accurate maps.
{"title":"ShinyDialect: a cartographic tool for spatial interpolation of geolinguistic data","authors":"C. Chagnaud, Philippe Garat, Paule-Annick Davoine, Elisabetta Carpitelli, Axel Vincent","doi":"10.1145/3149858.3149864","DOIUrl":"https://doi.org/10.1145/3149858.3149864","url":null,"abstract":"For decades, geolinguists have been using cartographic materials to display their data and understand the spatial distribution of local dialects. They have used many forms of maps such as maps with labels, symbols or colors. The most widely used are maps with isoglosses which are hand-drawn boundaries defining areas where people share the same language feature. Our source data is a mesh of georeferenced survey points representing a region of interest and each point of the mesh has a phonetic form attached. The issue is to transform this survey point mesh into homogeneous areas sharing similar categorical values in order to identify spatial patterns. This paper describes interpolation methods implemented to produce isogloss maps, namely turning a sample of points into areas boundaries. Implementing such methods requires performing spatial interpolations on a qualitative data set. We also describe a new cartographic tool, ShinyDialect, made for geolinguists to automate the construction process of maps with isoglosses in order to use it as support for spatial analysis. First, we will discuss geolinguistic data and the limits of the existing methods to compute linguistic areas. Next, we will describe the spatial interpolation methods we have implemented. Lastly, we will present features of the tool ShinyDialect to help geolinguists to build accurate maps.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"32 4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80134274","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Historians investigating evidence with spatial significance increasingly rely on gazetteers to identify the location of geographical features/places. Existing digital gazetteers cater to twenty-first century or discrete historical geographies (the classical world, for example). Early modernists (ca. 1450-1750), particularly those who work on non-Anglophone cultures, represent a major scholarly community with no temporally-appropriate gazetteers available. In this paper, we introduce a project that fills this research infrastructure gap. Mapping place names in the canonical eighteenth-century Encyclopédie is a case study for semi-automating the identification, classification, and location of places and spatial relations in historical geographic reference works printed in French. We demonstrate the challenges of using existing geoparsers and introduce our plan for new tools and protocols for working with historical French texts.
{"title":"Mapping the Encyclopédie: Working Towards an Early Modern Digital Gazetteer","authors":"Katherine McDonough, M. V. D. Camp","doi":"10.1145/3149858.3149861","DOIUrl":"https://doi.org/10.1145/3149858.3149861","url":null,"abstract":"Historians investigating evidence with spatial significance increasingly rely on gazetteers to identify the location of geographical features/places. Existing digital gazetteers cater to twenty-first century or discrete historical geographies (the classical world, for example). Early modernists (ca. 1450-1750), particularly those who work on non-Anglophone cultures, represent a major scholarly community with no temporally-appropriate gazetteers available. In this paper, we introduce a project that fills this research infrastructure gap. Mapping place names in the canonical eighteenth-century Encyclopédie is a case study for semi-automating the identification, classification, and location of places and spatial relations in historical geographic reference works printed in French. We demonstrate the challenges of using existing geoparsers and introduce our plan for new tools and protocols for working with historical French texts.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"37 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84435048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Violence and crime have been regarded as one of the notorious behaviors against humanity. With the rapid development of Information and Communications Technology (ICT), increasing amount of crime data become much more available and useful not only for police dispatch and crime prevention, but also for providing important references for the personal safety of local residents and visitors, especially in large cities. In this paper, we apply statistical approaches and graph theory to characterize the spatiotemporal properties of Chicago crime data from 2001 to 2016. First, we improved the previous Space-Time Kernel Density Estimation (STKDE) methods in computational efficiency. We proved that our improved method to compute STKDE has linear time computational complexity, which is experimentally verified to be much faster than previous methods. Second, we applied our improved STKDE method to demonstrate the intensities and hot spots of crime distribution in Chicago from 2001 to 2016. In order to reveal the displacement of crime incidents (i.e. movements of the hot spots), we detected the locations of highest crime hot spots at specified time intervals, and created hot spot displacement graphs based on whether a geographic location continues to be a crime hot spot across time intervals. Finally, the method of longest path on Directed Acyclic Graphs (DAG) was applied on the hot spot displacement graph in addition to the analysis of the number of components and their sizes of the graph. The result showed spatial crime displacement and temporal crime duration patterns. The proposed method advanced our knowledge in digital humanities, which can be applied to other cities, providing useful information for public safety.
{"title":"Disentangle crime hot spots and displacements in space and time: an analysis for Chicago from 2001 to 2016","authors":"Kai Wang, Xiaolu Zhou, Lixin Li","doi":"10.1145/3149858.3149860","DOIUrl":"https://doi.org/10.1145/3149858.3149860","url":null,"abstract":"Violence and crime have been regarded as one of the notorious behaviors against humanity. With the rapid development of Information and Communications Technology (ICT), increasing amount of crime data become much more available and useful not only for police dispatch and crime prevention, but also for providing important references for the personal safety of local residents and visitors, especially in large cities. In this paper, we apply statistical approaches and graph theory to characterize the spatiotemporal properties of Chicago crime data from 2001 to 2016. First, we improved the previous Space-Time Kernel Density Estimation (STKDE) methods in computational efficiency. We proved that our improved method to compute STKDE has linear time computational complexity, which is experimentally verified to be much faster than previous methods. Second, we applied our improved STKDE method to demonstrate the intensities and hot spots of crime distribution in Chicago from 2001 to 2016. In order to reveal the displacement of crime incidents (i.e. movements of the hot spots), we detected the locations of highest crime hot spots at specified time intervals, and created hot spot displacement graphs based on whether a geographic location continues to be a crime hot spot across time intervals. Finally, the method of longest path on Directed Acyclic Graphs (DAG) was applied on the hot spot displacement graph in addition to the analysis of the number of components and their sizes of the graph. The result showed spatial crime displacement and temporal crime duration patterns. The proposed method advanced our knowledge in digital humanities, which can be applied to other cities, providing useful information for public safety.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"194 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74821727","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Emotions influence people's behavior in a profound way. Feelings like happiness, hope, fear, boredom, anger, anxiety or relaxation affect the way people behave and interact with one another. However, there is often a strong correlation between the environment and the way people feel, e.g., the emotions associated with a hospital are typically very different from those associated with an amusement park or a promenade. The aim of an emotion map is to represent and depict interrelationships between emotions and geographic locations. Such maps can provide answers to various questions about how people feel at various places or at different times of the day. They can facilitate a search for places where people express a certain emotion. In this paper, we introduce a new approach of creating emotion maps from a large collection of geotagged social-media posts. We discuss potential usages of such maps. We present a model to query and utilize emotion maps and we demonstrate creation of emotion maps by applying emotion analysis to millions of geotagged tweets.
{"title":"Emotion Maps based on Geotagged Posts in the Social Media","authors":"Y. Doytsher, Ben Galon, Y. Kanza","doi":"10.1145/3149858.3149862","DOIUrl":"https://doi.org/10.1145/3149858.3149862","url":null,"abstract":"Emotions influence people's behavior in a profound way. Feelings like happiness, hope, fear, boredom, anger, anxiety or relaxation affect the way people behave and interact with one another. However, there is often a strong correlation between the environment and the way people feel, e.g., the emotions associated with a hospital are typically very different from those associated with an amusement park or a promenade. The aim of an emotion map is to represent and depict interrelationships between emotions and geographic locations. Such maps can provide answers to various questions about how people feel at various places or at different times of the day. They can facilitate a search for places where people express a certain emotion. In this paper, we introduce a new approach of creating emotion maps from a large collection of geotagged social-media posts. We discuss potential usages of such maps. We present a model to query and utilize emotion maps and we demonstrate creation of emotion maps by applying emotion analysis to millions of geotagged tweets.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"42 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82497741","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The digital geohumanities-and geographic computation generally-have advanced greatly by representing phenomena within geographic coordinate systems. More specifically, most visualizations and analyses only proceed once data are rendered into a single coordinate system via geolocation and one or more projections. But does it follow that geographic computation should require all phenomena to be represented in Euclidean or spherical geometry in a singular, absolute, Newtonian space? We suggest an approach to pluralizing the spaces available to geographic computation. We both supplement the technical architecture for projections and subtly reframe the purpose and meaning of projections. What we term numerical, generalized projections thereby become more central to GISystems. We suggest how existing libraries might be modified with minimal disruption (taking the widespread and foundational proj.4 library as example). We also envision modifications to existing OGC technical specifications for projections and coordinate systems. Finally, in conversation with the interpretative practice and nuanced spatialities of the digital geohumanities and critical geography, we further extend generalized projections to encompass spatial multiplicity, fragmented spaces, wormholes, and an expanded role for interruptions. This will facilitate: 1) interpretative approaches to scholarship and diverse constructions of space common in the humanities; 2) computational engagement with the ontological and epistemological commitments to relational space of critical human geography; and 3) scientific efforts to understand complex systems in the spaces and times that emerge from those systems' dynamics, revisiting a desire common in early quantitative geography; and 4) the desire for a broad basis of understanding geographic information in GIScience.
{"title":"Computing with many spaces: Generalizing projections for the digital geohumanities and GIScience.","authors":"Luke R Bergmann, David O'Sullivan","doi":"","DOIUrl":"","url":null,"abstract":"<p><p>The digital geohumanities-and geographic computation generally-have advanced greatly by representing phenomena within geographic coordinate systems. More specifically, most visualizations and analyses only proceed once data are rendered into a single coordinate system via geolocation and one or more projections. But does it follow that geographic computation should require all phenomena to be represented in Euclidean or spherical geometry in a singular, absolute, Newtonian space? We suggest an approach to pluralizing <i>the spaces</i> available to geographic computation. We both supplement the technical architecture for projections and subtly reframe the purpose and meaning of projections. What we term numerical, <i>generalized projections</i> thereby become more central to GISystems. We suggest how existing libraries might be modified with minimal disruption (taking the widespread and foundational proj.4 library as example). We also envision modifications to existing OGC technical specifications for projections and coordinate systems. Finally, in conversation with the interpretative practice and nuanced spatialities of the digital geohumanities and critical geography, we further extend generalized projections to encompass spatial multiplicity, fragmented spaces, wormholes, and an expanded role for interruptions. This will facilitate: 1) interpretative approaches to scholarship and diverse constructions of space common in the humanities; 2) computational engagement with the ontological and epistemological commitments to relational space of critical human geography; and 3) scientific efforts to understand complex systems in the spaces and times that emerge from those systems' dynamics, revisiting a desire common in early quantitative geography; and 4) the desire for a broad basis of understanding geographic information in GIScience.</p>","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"2017 ","pages":"31-38"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8188891/pdf/nihms-1591115.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"39100993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The digital geohumanities---and geographic computation generally--- have advanced greatly by representing phenomena within geographic coordinate systems. More specifically, most visualizations and analyses only proceed once data are rendered into a single coordinate system via geolocation and one or more projections. But does it follow that geographic computation should require all phenomena to be represented in Euclidean or spherical geometry in a singular, absolute, Newtonian space? We suggest an approach to pluralizing the spaces available to geographic computation. We both supplement the technical architecture for projections and subtly reframe the purpose and meaning of projections. What we term numerical, generalized projections thereby become more central to GISystems. We suggest how existing libraries might be modified with minimal disruption (taking the widespread and foundational proj.4 library as example). We also envision modifications to existing OGC technical specifications for projections and coordinate systems. Finally, in conversation with the interpretative practice and nuanced spatialities of the digital geohumanities and critical geography, we further extend generalized projections to encompass spatial multiplicity, fragmented spaces, wormholes, and an expanded role for interruptions. This will facilitate: 1) interpretative approaches to scholarship and diverse constructions of space common in the humanities; 2) computational engagement with the ontological and epistemological commitments to relational space of critical human geography; and 3) scientific efforts to understand complex systems in the spaces and times that emerge from those systems' dynamics, revisiting a desire common in early quantitative geography; and 4) the desire for a broad basis of understanding geographic information in GIScience.
{"title":"Computing with many spaces: Generalizing projections for the digital geohumanities and GIScience","authors":"Luke Bergmann, David O'Sullivan","doi":"10.1145/3149858.3149866","DOIUrl":"https://doi.org/10.1145/3149858.3149866","url":null,"abstract":"The digital geohumanities---and geographic computation generally--- have advanced greatly by representing phenomena within geographic coordinate systems. More specifically, most visualizations and analyses only proceed once data are rendered into a single coordinate system via geolocation and one or more projections. But does it follow that geographic computation should require all phenomena to be represented in Euclidean or spherical geometry in a singular, absolute, Newtonian space? We suggest an approach to pluralizing the spaces available to geographic computation. We both supplement the technical architecture for projections and subtly reframe the purpose and meaning of projections. What we term numerical, generalized projections thereby become more central to GISystems. We suggest how existing libraries might be modified with minimal disruption (taking the widespread and foundational proj.4 library as example). We also envision modifications to existing OGC technical specifications for projections and coordinate systems. Finally, in conversation with the interpretative practice and nuanced spatialities of the digital geohumanities and critical geography, we further extend generalized projections to encompass spatial multiplicity, fragmented spaces, wormholes, and an expanded role for interruptions. This will facilitate: 1) interpretative approaches to scholarship and diverse constructions of space common in the humanities; 2) computational engagement with the ontological and epistemological commitments to relational space of critical human geography; and 3) scientific efforts to understand complex systems in the spaces and times that emerge from those systems' dynamics, revisiting a desire common in early quantitative geography; and 4) the desire for a broad basis of understanding geographic information in GIScience.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91387093","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Knowing where people live is a fundamental component of many decision making processes such as urban development, infectious disease containment, evacuation planning, risk management, conservation planning, and more. While bottom-up, survey driven censuses can provide a comprehensive view into the population landscape of a country, they are expensive to realize, are infrequently performed, and only provide population counts over broad areas. Population disaggregation techniques and population projection methods individually address these shortcomings, but also have shortcomings of their own. To jointly answer the questions of "where do people live" and "how many people live there," we propose a deep learning model for creating high-resolution population estimations from satellite imagery. Specifically, we train convolutional neural networks to predict population in the USA at a 0.01°x0.01° resolution grid from 1-year composite Landsat imagery. We validate these models in two ways: quantitatively, by comparing our model's grid cell estimates aggregated at a county-level to several US Census county-level population projections, and qualitatively, by directly interpreting the model's predictions in terms of the satellite image inputs. We find that aggregating our model's estimates gives comparable results to the Census county-level population projections and that the predictions made by our model can be directly interpreted, which give it advantages over traditional population disaggregation methods. In general, our model is an example of how machine learning techniques can be an effective tool for extracting information from inherently unstructured, remotely sensed data to provide effective solutions to social problems.
{"title":"A Deep Learning Approach for Population Estimation from Satellite Imagery","authors":"Caleb Robinson, Fred Hohman, B. Dilkina","doi":"10.1145/3149858.3149863","DOIUrl":"https://doi.org/10.1145/3149858.3149863","url":null,"abstract":"Knowing where people live is a fundamental component of many decision making processes such as urban development, infectious disease containment, evacuation planning, risk management, conservation planning, and more. While bottom-up, survey driven censuses can provide a comprehensive view into the population landscape of a country, they are expensive to realize, are infrequently performed, and only provide population counts over broad areas. Population disaggregation techniques and population projection methods individually address these shortcomings, but also have shortcomings of their own. To jointly answer the questions of \"where do people live\" and \"how many people live there,\" we propose a deep learning model for creating high-resolution population estimations from satellite imagery. Specifically, we train convolutional neural networks to predict population in the USA at a 0.01°x0.01° resolution grid from 1-year composite Landsat imagery. We validate these models in two ways: quantitatively, by comparing our model's grid cell estimates aggregated at a county-level to several US Census county-level population projections, and qualitatively, by directly interpreting the model's predictions in terms of the satellite image inputs. We find that aggregating our model's estimates gives comparable results to the Census county-level population projections and that the predictions made by our model can be directly interpreted, which give it advantages over traditional population disaggregation methods. In general, our model is an example of how machine learning techniques can be an effective tool for extracting information from inherently unstructured, remotely sensed data to provide effective solutions to social problems.","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"13 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81740987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities","authors":"","doi":"10.1145/3149858","DOIUrl":"https://doi.org/10.1145/3149858","url":null,"abstract":"","PeriodicalId":93223,"journal":{"name":"Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)","volume":"181 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2017-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80211625","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM SIGSPATIAL Workshop on Geospatial Humanities (1st : 2017 : Redondo Beach, Calif.)
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：