[Back]

M. Tosic:
"Analysing the Potential of Network Kernel Density Estimation for the Study of Tourism Based on Geosocial Media Data";
Supervisor: F. Porras Bernárdez, N. van de Weghe, G. Gartner; Department für Geodäsie und Geoinformation, FB Kartographie, 2019; final examination: 2019-09-26.

Social media today has an important role in promoting tourist destinations. It is considered as a useful and reliable source of tourist information. The analysis of its big data can give an important insight into tourists´ behaviour and preferences. By using geotagged photos from social media sites, e.g. Flickr, destination management organisations can predict tourist behaviour and patterns at a destination. A major challenge today is how to track these behavioural patterns. Some of the possible methods are Kernel Density Estimation and Network Kernel Density Estimation. In this research, both analyses were used to identify areas and streets of interest of Brussels by using Flickr dataset. The assumption is that using NKDE in tourism could lead to the identification of the most popular tourist places in a city. NKDE is a density analysis on a network and it was used to identify the most popular tourist street segments in a city, while KDE was used to determine tourist areas of interest. To define tourist places, it was necessary to acquire tourist information about landmarks from various sources and compare them with the results of analyses. Areas and street segments of interest were also determined for visitors from different countries of origin. All countries of origin were included in the research and they were categorized in groups. The final product is maps presenting integration of results from both density analyses and maps with tourist attractions, per each group. Both KDE and NKDE were compared and evaluated, main differences between results of the analyses were established, and advantages and disadvantages of both methods were defined.

Kernel Density Estimation, Network Kernel Density Estimation, Flickr, Geosocial Media, Tourism, Map visualization, Areas of Interest, Street Segment of Interest