Rising mental health issues among adolescents have increased interest in automated approaches for detecting early signs of psychological distress in digital text. One important focus is the identification of cognitive distortions – irrational thought patterns – because of their role in aggravating mental distress, and early detection may enable timely, low cost interventions. While prior work has focused on English data, we present a first in-depth study of cross lingual and cross register generalization for cognitive distortion detection, using forum posts written by Dutch adolescents. We frame the task at two levels: (1) detecting whether a post contains a cognitive distortion, and (2) identifying the specific text span that expresses it. Our findings show that domain adaptation methods perform best for post-level detection, while a simpler technique – sentence embeddings with a classifier – outperforms more complex models for span identification. Results show predicting cognitive distortions in text is challenging, and highlight how changes in language and writing style can significantly impact performance.

Today, huge volumes of information flow at unprecedented speeds, and large organisations like the Dutch National Police face significant challenges. Efficiently sharing, processing, and prioritizing information is essential but has become increasingly difficult to achieve. These chal- lenges often hinder their ability to make sound and timely decisions. This study investigates innovative methods to enhance collaborative decision-making within such complex environ- ments.
The main goal of this study is to research and test the potential of Artificial Swarming Intelli- gence to enable collective ranking of information based on its importance. The resulting Proof of Concept offers a new method for improving the efficiency of information sharing, enhancing collective intelligence, and facilitating the decision-making processes within the police force.
Drawing inspiration from natural swarming behaviour seen in species like bees, the platform allows participants to rank multiple pieces of information during collaborative sessions. This process helps to highlight the most significant topics, enabling quicker access to critical in- sights.
The Proof of Concept is structured as a digital platform designed to improve real-time decision- making. For our case study, we focused on the policy advisors of the Dutch National Police. The Proof of Concept operates as a client-server model, ensuring that user interactions are efficient while providing live updates on rankings. This allows participants to engage actively and see the evolving importance of various pieces of information.
However, despite the progress achieved, several limitations became evident. First, some envisioned functionalities were not fully developed or implemented within the project’s time frame. In addition, current design faces challenges in scalability, particularly when engaging larger groups of users. Issues related to user accessibility and potential biases in decision- making processes also came into view. To guide further work, we have highlighted key areas for future research and provided specific considerations that address these limitations, aiming to support the Proof of Concept’s development into a complete and fully scalable platform.
Ultimately, this Proof of Concept presents a promising approach to enhancing decision-making processes at the Dutch National Police. It lays the foundation for future research and devel- opment, with opportunities to refine its functionality and expand its use within the organisation. By improving how information is shared and prioritised, the Dutch National Police can enhance its collective intelligence and improve its decision-making processes.