Spatial analyses of food safety, customer ratings and price of restaurants in Denmark
Author
Baescu, Oana
Term
4. term
Publication year
2017
Abstract
This thesis investigates whether relationships exist between the food safety status of Danish restaurants (cleanliness assessed via the Danish Food Administration’s Smiley scheme), their price levels, and the customer ratings they receive, viewed through a spatial lens. The dataset comprises 1,492 restaurants across Denmark, including 778 in the Copenhagen area, assembled by validating records from the Smiley database and retrieving price, category, and rating information from Foursquare via its API in Python. The combined data are analyzed and visualized in ArcMap, QGIS, and R using methods such as kernel density estimation, Ripley’s K-function, Voronoi diagrams, and choropleth mapping, and by exploring Spearman correlations between inspection variables (e.g., elite smiley and last inspection) and ratings and price. Analyses are conducted at two scales (national and Copenhagen) to map distributions, identify clustering, and examine potential associations. The excerpt outlines the workflow and outputs (maps and correlation plots) but does not provide specific numerical findings; results and limitations are discussed in the full thesis.
Dette speciale undersøger, om der findes sammenhænge mellem danske restauranters fødevaresikkerhed (renlighed vurderet gennem Fødevarestyrelsens Smiley-ordning), deres prisniveau og de kundevurderinger, de modtager, i et rumligt perspektiv. Datagrundlaget omfatter 1.492 restauranter i Danmark, heraf 778 i Københavnsområdet, og er sammensat ved at validere poster fra Smiley-databasen og hente pris-, kategori- og ratingoplysninger fra Foursquare via API’et i Python. De kombinerede data analyseres og visualiseres i ArcMap, QGIS og R med metoder som kerne-tæthetsestimering, Ripley’s K-funktion, Voronoi-diagrammer og koroplethkort, samt Spearman-korrelationer mellem inspektionsvariabler (fx elite-smiley og sidste kontrol) og ratings og pris. Analyserne gennemføres på to skalaer (nationalt og København) for at kortlægge fordeling, identificere klynger og undersøge potentielle sammenhænge. Uddraget skitserer arbejdsprocessen og output (kort og korrelationsplots), men indeholder ikke specifikke numeriske resultater; resultater og begrænsninger drøftes i den fulde afhandling.
[This apstract has been generated with the help of AI directly from the project full text]