Centre de documentation
scientifique

Deep learning method for Chinese multisource point of interest matching / Pengpeng Li in Computers, Environment and Urban Systems, vol 96 (September 2022)  

Public [article]  inComputers, Environment and Urban Systems > vol 96 (September 2022) . - n° 101821 Titre : Deep learning method for Chinese multisource point of interest matching Type de document : Article/Communication Auteurs : Pengpeng Li, Auteur ; Jiping Liu, Auteur ; An Luo, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 101821 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique [Termes IGN] appariement sémantique [Termes IGN] apprentissage profond [Termes IGN] classification par Perceptron multicouche [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] extraction de traits caractéristiques [Termes IGN] inférence sémantique [Termes IGN] information sémantique [Termes IGN] point d'intérêt [Termes IGN] représentation vectorielle [Termes IGN] traitement du langage naturel Résumé : (auteur) Multisource point of interest (POI) matching refers to the pairing of POIs that refer to the same geographic entity in different data sources. This also constitutes the core issue in geospatial data fusion and update. The existing methods cannot effectively capture the complex semantic information from a text, and the manually defined rules largely affect matching results. This study developed a multisource POI matching method based on deep learning that transforms the POI pair matching problem into a binary classification problem. First, we used three different Chinese word segmentation methods to segment the POI text attributes and used the segmentation results to train the Word2Vec model to generate the corresponding word vector representation. Then, we used the text convolutional neural network (Text-CNN) and multilayer perceptron (MLP) to extract the POI attributes' features and generate the corresponding feature vector representation. Finally, we used the enhanced sequential inference model (ESIM) to perform local inference and inference combination on each attribute to realize the classification of POI pairs. We used the POI dataset containing Baidu Map, Tencent Map, and Gaode Map from Chengdu to train, verify, and test the model. The experimental results show that the matching precision, recall rate, and F1 score of the proposed method exceed 98% on the test set, and it is significantly better than the existing matching methods. Numéro de notice : A2022-513 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/INFORMATIQUE Nature : Article DOI : 10.1016/j.compenvurbsys.2022.101821 Date de publication en ligne : 18/06/2022 En ligne : https://doi.org/10.1016/j.compenvurbsys.2022.101821 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101053 [article]

A hybrid classification matching method for geospatial services / Yandong Wang in Transactions in GIS, vol 16 n° 6 (December 2012)  

Public [article]  inTransactions in GIS > vol 16 n° 6 (December 2012) . - pp 781 - 805 Titre : A hybrid classification matching method for geospatial services Type de document : Article/Communication Auteurs : Yandong Wang, Auteur ; Hao Li, Auteur ; An Luo, Auteur Année de publication : 2012 Article en page(s) : pp 781 - 805 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géomatique web [Termes IGN] classification hybride [Termes IGN] service web géographique [Termes IGN] système d'information géographique [Termes IGN] taxinomie Résumé : (Auteur) With the development of the Internet and GIS, large volumes of spatial data and powerful computing resources are increasingly published in the form of Web services. Given the variety and number of geospatial services advertised online, finding appropriate geospatial services has become a tremendous challenge for potential users. Geospatial service classification provides a basis for developing matching criteria to improve the efficiency of service discovery. At present, most classification-based matching methods require users to provide classification descriptions using a specified taxonomy. These requirements seriously limit the application of classification-based matching. To solve these kinds of problems, this article presents a hybrid geospatial service classification-matching method. Based on the differences in classification descriptions, three strategies are proposed: (1) the existing classification matching method is used for requests with classifications described using homogeneous taxonomies; (2) a formal-concept-analysis-based service classification matching method is proposed for service requests with classifications described using heterogeneous taxonomies; and (3) an interface-similarity-based service classification decision method is proposed for requests without classification descriptions. The feasibility of the hybrid geospatial service classification matching method is verified by two sets of experiments. The results reveal that this method can effectively broaden the application of classification-based matching in geospatial service discovery. Numéro de notice : A2012-616 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1111/j.1467-9671.2012.01348.x En ligne : https://doi.org/10.1111/j.1467-9671.2012.01348.x Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=32062 [article]