Diploma and Master Theses (authored and supervised):

M. Ibrahim:
"Extracting and Mapping Spatial Patterns of Interest from Social Media";
Supervisor: G. Gartner, F. Porras Bernárdez; Department für Geodäsie und Geoinformation, FB Kartographie, 2020; final examination: 2020-01-16.

English abstract:
Spatial patterns of interest are those spatial extents of interest to people. These patterns can be identified in two forms. The first form is areas of interest (AOIs) which refers to those areas that attract peoplesī interest within the city borders. The spatial extent of these areas is not defined because AOIs are subjective to the behavior of different groups of people. The second form of visualizing spatial patterns of interest is footprints which refers to spatial spread of human activity.
Studying and understanding spatial patterns of interest, would undoubtedly help us to understand the spatial behavior of different groups of people. Thus, the extracted information about their patterns can be very useful in a lot of applications such as urban planning, location based services, national security, tourism, transportation and much more. Geosocial media GSM as stores different types of data such as tags, comments and photos that correlates to a space. Thus, GSM as a data source has proven reliability to provide data for spatial pattern analysis.
This thesis introduces a work scheme to extract and analyze spatial patterns of interest of different groups of tourists in Vienna. The proposed work paradigm can be used to identify spatial patterns of interest in similar cases. Vienna has been chosen as the study area in this research. Vienna has special attributes that make the city a standard area for touristic pattern analysis such as history, culture, location, event hosting, political situation and climate.
A dataset of Flickr geotagged photo data has been used to analyze the interest patterns in Vienna. Flickr is one of the oldest photo sharing platforms, thus Flickr data is very useful for performing a spatio-temporal analysis to compare the density evolution of touristsī activities during a wide time period than what other GSM data would provide.
A density based clustering algorithm HDBSCAN has been used to identify AOIs from Flickr data with help of concave hull enveloping algorithm. Different tourist groupsī footprints have been identified using Kernel density estimation KDE. Both the followed methodology have proven reliability to identify interest patterns of tourists in Vienna. Using two different methodologies in this case study, has opened a door to criticize both methodologies during the research.

Created from the Publication Database of the Vienna University of Technology.