ECIS 2020 Research Papers


Internet users typically do not read the privacy policies of websites since they are written as complex legal texts and at great length. Recent research addresses this issue by applying Natural Language Processing and Machine Learning approaches in order to strengthen the digital sovereignty of Internet users by extraction of relevant information of privacy policies. These approaches achieve an accuracy of up to 90%. However, none of these have successfully prevailed, due to insufficient consideration of the requirements of especially privacy-aware Internet users. Therefore, we present the architecture of AMARYLLIS (AutoMAted pRivacY poLicy anaLysIS), a user-centric information system, as well as its use cases, applying a Design Science Research methodology. Our information system maps the entire privacy policy analysis process against users’ preferences, provides a scalable solution, incorporates a usercentric design, and includes Privacy by Design. An evaluation with potential users and experts reveals significant satisfaction with these features. The results highlight the importance of features currently not considered by existing solutions. Therefore, these features should serve as fundamental design principles of an information system analyzing privacy policies.



When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.