Deanonymizace otevřených dat o kriminalitě

The thesis deals with the topic of crime open data deanonymization. Many government organizations publish open data or are at least considering it. Before publication, the data is anonymized to prevent sensitive information leaks. Nevertheless, various methods can reverse this process and recover sensitive information about the data owner. Objectives of this paper have been set regarding this issue. The first objective was to document the current state of crime open data publication in Australia, the European Union, Canada, New Zealand, the United Kingdom of Great Britain and Northern Ireland, and the United States of America. Research and analysis of available sources were used to fulfill this objective. The second objective was to analyze the different ways and methods for data deanonymization. Analysis of available literature was used to meet this objective. The final objective was to perform deanonymization on a real-world dataset. A suitable dataset was selected, which comes from the Avon and Somerset area in the United Kingdom and covers crime data from March 2021, as well as related sources to find the possible owners of the crime data - Bristol Live and Somerset Live. The discussion at the end of the paper goes into the difference between manual and automated approaches to deanonymization. The results were also put in the context of the current state of knowledge. The contributions of this thesis are to provide insight into the crime open data publications around the world, describe a few selected methods of deanonymization, and finally, give an example of re-identification done on a real-world dataset.