In this paper we use the gravity model to estimate the similarity of US cities based on data provided by Google Trends (GT). GT allows to look up search terms and to obtain ranked lists of US cities according to the relative frequencies of requests for each term. The occurences of the US cities on these ranked lists are used to determine the similarities with the gravity model. As search terms for GT serve dictionaries derived from the General Inquirer (GI), containing the categories Economy and Politics/Legal. The estimated similarity scores are visualized with multidimensional scaling (MDS).
Anzeige
Bitte loggen Sie sich ein, um Zugang zu diesem Inhalt zu erhalten
Note that the gravity distance is not a distance measure in the strict mathematical sense, since it neither fulfills d(x, x) = 0 nor the triangle inequality.