Automated Hyperlink Text Analysis of City Websites: Projected Image Representation on the Web

Abstract

The objective of this study is to identify the image representations of 75 European cities on the Web. As an effective image positioning strategy will result in successful differentiation from competitors, it is crucial for tourism destinations to regularly examine their image. This study focuses on the supply-side of destination image formation and is therefore concerned with analysing the projected destination image. Hyperlink text of DMO websites was collected automatically by a crawler. The text was then edited and filtered. Latent semantic dimensions were generated by applying PCA. A hierarchical cluster approach revealed different groups of hyperlink terms. Finally, the co-occurrence of terms and cities was displayed in a joint-map indicating which groups of hyperlink terms are over- or underrepresented for each of the cities. This information permits conclusions regarding the projected image of the cities.