Nowadays there are a lot of geographic internet services that allow the search and display of geographic maps. While most of them are owned and developed by companies, with the diffusion of online crowd-sourced initiative, other providers started to develop open-source collaborative projects. In this context, OpenStreetMap is one of the most famous and used platform providing free, editable maps and cartographies of the World. Due to its open nature, this platform has been subject of interest for researchers all over the world inspiring analysis and studies about different areas in the planet and about people habits. Indeed, in addition to the fact that its data is free and obtainable by anyone, the data it provides about the environment and the places is deemed to be less biased, because it is crowd-sourced and not defined by private companies. This thesis work takes inspiration from the multitude of research studies carried out leveraging OpenStreetMap and exploits its data in order to understand what impact a phenomenon like the Covid-19 pandemic had on cities and their people. To do so, the work described in this thesis started with a data analytics task to understand the main trends and what aspects of the data could be further analyzed to identify patterns that could be related to the virus widespread and its impact on the society. The analysis was first performed through a visual approach, using statistical tests to confirm the significance of the observed variations. The second part of this thesis, instead, describes a more advanced analysis campaign carried out by leveraging ad-hoc machine learning models. These models were exploited to understand to what extent the evolution of the maps recorded during 2020 were expected or can be linked to changes in people habits caused by Covid-19.
Comprendere l'impatto del Covid-19 sull'interazione delle persone con le città: un'analisi di Open Street Map. Al giorno d'oggi esistono moltissimi servizi geografici online che consentono la ricerca e la visualizzazione di mappe geografiche. Mentre la maggior parte di essi sono di proprietà e sviluppati da aziende, con la diffusione di iniziative di crowdsourcing online, altri provider hanno iniziato a sviluppare progetti collaborativi open source. In questo contesto, OpenStreetMap è una delle piattaforme più famose e utilizzate che fornisce mappe e cartografie gratuite e modificabili del mondo. Per la sua natura aperta, questa piattaforma è stata oggetto di interesse per i ricercatori di tutto il mondo, ispirando analisi e studi sulle diverse aree del pianeta e sulle abitudini delle persone. Infatti, oltre al fatto che i suoi dati sono gratuiti e ottenibili da chiunque, i dati che fornisce sull'ambiente e sui luoghi sono ritenuti più imparziali, perché crowd-sourced e non definiti da aziende private. Questo lavoro di tesi prende ispirazione dalla moltitudine di studi di ricerca condotti sfruttando OpenStreetMap e sfrutta i suoi dati per capire quale impatto abbia avuto un fenomeno come la pandemia causata dal Covid-19 sulle città e sulla loro gente. Per fare ciò, il lavoro descritto in questa tesi è partito con una analisi dei dati iniziale per comprendere le principali tendenze e quali aspetti dei dati potrebbero essere ulteriormente analizzati per identificare modelli che potrebbero essere correlati al virus diffuso e al suo impatto sulla società. L'analisi è stata prima eseguita attraverso un approccio visivo, utilizzando test statistici per confermare la significatività delle variazioni osservate. La seconda parte di questa tesi, invece, descrive una campagna di analisi più avanzata realizzata sfruttando modelli di machine learning ad-hoc. Questi modelli sono stati sfruttati per capire fino a che punto l'evoluzione delle mappe registrate nel corso del 2020 fosse prevista o collegabile ai cambiamenti nelle abitudini delle persone causati dal Covid-19.
Understanding the impact of Covid-19 on the interaction of people with cities: An analysis of Open Street Map.
LIETAVEC, DOMINIK
2020/2021
Abstract
Nowadays there are a lot of geographic internet services that allow the search and display of geographic maps. While most of them are owned and developed by companies, with the diffusion of online crowd-sourced initiative, other providers started to develop open-source collaborative projects. In this context, OpenStreetMap is one of the most famous and used platform providing free, editable maps and cartographies of the World. Due to its open nature, this platform has been subject of interest for researchers all over the world inspiring analysis and studies about different areas in the planet and about people habits. Indeed, in addition to the fact that its data is free and obtainable by anyone, the data it provides about the environment and the places is deemed to be less biased, because it is crowd-sourced and not defined by private companies. This thesis work takes inspiration from the multitude of research studies carried out leveraging OpenStreetMap and exploits its data in order to understand what impact a phenomenon like the Covid-19 pandemic had on cities and their people. To do so, the work described in this thesis started with a data analytics task to understand the main trends and what aspects of the data could be further analyzed to identify patterns that could be related to the virus widespread and its impact on the society. The analysis was first performed through a visual approach, using statistical tests to confirm the significance of the observed variations. The second part of this thesis, instead, describes a more advanced analysis campaign carried out by leveraging ad-hoc machine learning models. These models were exploited to understand to what extent the evolution of the maps recorded during 2020 were expected or can be linked to changes in people habits caused by Covid-19.È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: unitesi@unipv.it.
https://hdl.handle.net/20.500.14239/13175