<?xml version='1.0' encoding='UTF-8'?><metadata xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:dcterms="http://purl.org/dc/terms/" xmlns="http://dublincore.org/documents/dcmi-terms/"><dcterms:title>CityMe - dataset with mapped regions</dcterms:title><dcterms:identifier>https://doi.org/10.82210/E5LK1Y</dcterms:identifier><dcterms:creator>Costa, Elvira</dcterms:creator><dcterms:publisher>Repositório Polen QTY</dcterms:publisher><dcterms:issued>2024-05-07</dcterms:issued><dcterms:modified>2024-05-07T16:20:54Z</dcterms:modified><dcterms:description>The main components of the framework include: setting the aggregated textual data from Twitter and place-based sources (Google and OSM); employing the BERTopic transformer-based topic modeling for each source; comparing topics emerged from each source using the cosine similarity metric; carrying out Getis-Ord G∗i
 hotspot analysis for retrieving statistically representative topic-based cells; applying the Jaccard similarity index aimed at ultimately comparing thematic and spatial similarities that support the discussion on content-location relationships for the case study.</dcterms:description><dcterms:subject>Social Sciences - Economic and Social Geography - Urban Studies</dcterms:subject><dcterms:isReferencedBy>Tang, V., &amp; Painho, M. (2023). Content-location relationships: a framework to explore correlations between space-based and place-based user-generated content. International Journal of Geographical Information Science, 37(8), 1840–1871. https://doi.org/10.1080/13658816.2023.2213869, doi, 10.1080/13658816.2023.2213869, https://doi.org/10.1080/13658816.2023.2213869</dcterms:isReferencedBy><dcterms:date>2024-05-07</dcterms:date><dcterms:contributor>Costa, Elvira</dcterms:contributor><dcterms:contributor>Leonardo Vanneschi</dcterms:contributor><dcterms:dateSubmitted>2024-05-07</dcterms:dateSubmitted><dcterms:license>CC0 1.0</dcterms:license></metadata>