Centre de documentation
scientifique

A machine learning approach for detecting rescue requests from social media / Zheye Wang in ISPRS International journal of geo-information, vol 11 n° 11 (November 2022)  

Public [article]  inISPRS International journal of geo-information > vol 11 n° 11 (November 2022) . - n° 570 Titre : A machine learning approach for detecting rescue requests from social media Type de document : Article/Communication Auteurs : Zheye Wang, Auteur ; Nina S.N. Lam, Auteur ; Mingxuan Sun, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 570 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique web [Termes IGN] apprentissage automatique [Termes IGN] code postal [Termes IGN] données d'entrainement (apprentissage automatique) [Termes IGN] données issues des réseaux sociaux [Termes IGN] Etats-Unis [Termes IGN] filtrage d'information [Termes IGN] secours d'urgence [Termes IGN] tempête [Termes IGN] terminologie [Termes IGN] Twitter Résumé : (auteur) Hurricane Harvey in 2017 marked an important transition where many disaster victims used social media rather than the overloaded 911 system to seek rescue. This article presents a machine-learning-based detector of rescue requests from Harvey-related Twitter messages, which differentiates itself from existing ones by accounting for the potential impacts of ZIP codes on both the preparation of training samples and the performance of different machine learning models. We investigate how the outcomes of our ZIP code filtering differ from those of a recent, comparable study in terms of generating training data for machine learning models. Following this, experiments are conducted to test how the existence of ZIP codes would affect the performance of machine learning models by simulating different percentages of ZIP-code-tagged positive samples. The findings show that (1) all machine learning classifiers except K-nearest neighbors and Naïve Bayes achieve state-of-the-art performance in detecting rescue requests from social media; (2) using ZIP code filtering could increase the effectiveness of gathering rescue requests for training machine learning models; (3) machine learning models are better able to identify rescue requests that are associated with ZIP codes. We thereby encourage every rescue-seeking victim to include ZIP codes when posting messages on social media. This study is a useful addition to the literature and can be helpful for first responders to rescue disaster victims more efficiently. Numéro de notice : A2022-846 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.3390/ijgi11110570 Date de publication en ligne : 16/11/2022 En ligne : https://doi.org/10.3390/ijgi11110570 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102081 [article]

Construction of the planar partition postal code map based on cadastral registration / F. Penninga in Geoinformatica, vol 9 n° 2 (June - August 2005)  

Public [article]  inGeoinformatica > vol 9 n° 2 (June - August 2005) . - pp 181 - 204 Titre : Construction of the planar partition postal code map based on cadastral registration Type de document : Article/Communication Auteurs : F. Penninga, Auteur ; E. Verbee, Auteur ; et al., Auteur Année de publication : 2005 Article en page(s) : pp 181 - 204 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Systèmes d'information géographique [Termes IGN] cadastre étranger [Termes IGN] cadastre numérique [Termes IGN] code postal [Termes IGN] diagramme de Voronoï [Termes IGN] géocodage par adresse postale [Termes IGN] Pays-Bas [Termes IGN] squelettisation [Termes IGN] World Geodetic System 1984 Résumé : (Auteur) Accurate postal code maps have many applications within GIS as the postal code has the potential to link the address description of buildings to their location in a specified global reference system in a more natural way. This relationship is possible in both directions: geocoding and reverse-geocoding. These operators demand a mechanism for translating an exact geometric position (i.e., WGS84 coordinate) into a location indication (town, street, house number) and vice versa. As most built-up parcels are provided with a postal code, this indicator can be used as the linkage. This paper describes the procedure, based on the Dutch cadastral registration, to obtain a reliable 6-position (i.e., 262813X, the highest level of detail possible) planar postal code map for the Netherlands. Problems with existing, Voronoi-diagram based, postal code maps, like intersected houses and arbitrary derived (and thus unrecognizable) boundaries are avoided. The reliability of the derived planar postal code map is discussed and results are illustrated by figures. For a planar coverage, non built-up parcels having no postal code should be assigned a plausible postal code. Furthermore special attention is given to infrastructural parcels. These parcels are divided at their (approximated) skeletons first and then these subdivided infrastructure parcels are piecewise attached to their neighbour parcels. This new approach results in very reliable postal code maps, which are visually attractive too as infrastructure lines can be recognized. The procedure is generic and can be applied to other administrative parcel information as well. The algorithm is implemented using the Computational Geometry Algorithms Library (CGAL), and the possibilities and limitations of this library are addressed as well. Also a number of non-implemented alternatives or improvements are given. Numéro de notice : A2005-226 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/INFORMATIQUE Nature : Article DOI : 10.1007/s10707-005-6430-3 En ligne : https://doi.org/10.1007/s10707-005-6430-3 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=27363 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
057-05021	RAB	Revue	Centre de documentation	En réserve L003	Disponible