Pavol Sokol

The increasing availability of legal documentation provides opportunities for applying empirical research methods that enable a comprehensive analysis of complex legal data. Systematic Content Analysis offers a structured approach for examining legal texts. This paper examines the possibility of anomaly detection in legal text through the employment of machine learning methods. The conclusions presented are formulated on a basis of a case study focusing on contractual agreements concluded between public authorities and private service providers in the field of cybersecurity. In this regard, we propose an approach to the detection of anomalous legal text, providing a categorization of attributes that impact the determination of anomalies and describing their significance in the selected domain.

Sharing the alerts from intrusion detection systems among multiple computer networks and organizations allows for seeing the “big picture” of the network security situation and improves the capabilities of cyber incident response. However, such a task requires a number of technical and non-technical issues to be resolved, from data collection and distribution to proper categorization, data quality management, and issues of trust and privacy. In this field note, we illustrate the concepts and provide lessons learned on the example of SABU, an alert sharing and analysis platform used by academia and partner organizations in the Czech Republic. We discuss the initial willingness to share the data that was later weakened by the uncertainties around personal data protection, the issues of high volume and low quality of the data that prevented their straightforward use, and that the management of the community is a more severe issue than the technical implementation of alert sharing.