· Hub for the NS Community ·
Dear community members,
From identifying phishing attempts in mails to breaking new ground in detecting violations of the NST theorems, the first AI Conference for Software Engineering proved to be an interesting gathering for promising AI tools.
In November NSX participated in The First International Conference on AI-based Software Engineering for Digital Services (AISED 2025). Speakers from many academic institutions and private organizations shared their research and findings on integrating AI.The most notable examples include using AI to recognize e-mail writing patterns to identify spear phishing attempts, research on refactoring code with LLMs, UI testing with RPA and AI and automatic identification of published advertisements related to human trafficking.
In addition to these talks, the conference hosted panels in which industry experts discussed related topics such as "Will AI take over our jobs" and "To what extent can we trust AI generated code".
NSX also shared their ongoing progress and hosted several presentations on advancing software factories, using LLMs to identify NST theorem violations and a NS approach towards creating a pipeline for research data analysis.
Vojtěch Knaisl and Robert Pergl published a new article in which they introduce a novel framework that advances data management planning (DMP) by integrating the principles of Normalized Systems Theory (NST) to enhance modularity, scalability, and machine-actionability. The authors demonstrate how semantic technologies, ontologies and linked data can transform traditional DMPs into interoperable, evolvable artefacts — enabling automated evaluation, simplified adaptation to new funding or policy templates, and stronger alignment with FAIR principles. Through a real-world case study, they illustrate how their approach supports researchers and data stewards by reducing manual effort, improving guidance quality, and strengthening sustainable data stewardship practices.
The work presented in this paper incorporates various earlier research results which are interesting to the NS community, a.o. the usage of linked data, generating questionnaires based on models to capture user input (in contrast to using CRUDS screens), transforming models based on evolving metamodels, and document generation. An interesting read!
Reference:
Knaisl, V., & Pergl, R. (2025). Machine-Actionability and Evolvability in Data Stewardship Planning:
Framework, Implementation, and Case Study. Data Science Journal, 24. DOI: w10.5334/dsj-2025-025
The iPSYcare project, which focuses on creating a standardized and evolvable data infrastructure for mental healthcare in Belgium, is approaching the end of its current phase. iPSYcare is a platform for the secondary use of data from five psychiatric hospitals in the Antwerp region, enabling research, quality monitoring and policy insights based on harmonized and validated information.
Throughout this phase, NSX delivered the full technical implementation: a validated FHIR Client, the iPSYcare data platform, and the Power BI dashboard that provides insights on top of it. The development was carried out in close collaboration with the participating psychiatric hospitals and the Belgian government, which funded the project. This ensured that the platform aligns with real clinical workflows and national interoperability requirements. The FHIR expanders discussed in an earlier newsletter were used to build the data platform, highlighting the domain-specific scalability of our approach.
iPSYcare has been developed together with the University of Antwerp, where the platform is also hosted. The initiative originated from and continues to be supported by the Public Mental Health Fund of the University of Antwerp, which aims to strengthen mental-health innovation through evidence-based, data-driven methods. This academic grounding ensures a secure, research-oriented environment and enables future studies and longitudinal analyses.
As the current phase concludes, iPSYcare will be extended, allowing the consortium to further expand the platform. As national guidelines and FHIR profiles continue to mature, the dataset will be extended with new variables, strengthening the analytical possibilities for researchers, clinicians and policymakers. Thanks to NSX's metamodel-driven approach, the platform can evolve without compromising stability.
The continuation of iPSYcare will further enhance interoperability, data quality and advanced analytics in the mental healthcare sector, enabling Belgium to build future improvements on a robust, evolvable digital infrastructure.
