Governments and public organisations around the globe collect, produce and (openly) disseminate vast amounts of data, including, for example, deliberation data, real-time sensor data, statistics, legal documents, meteorological data, satellite imagery, environmental data, and electronic health records. The expected benefits of these data for citizens, businesses, public administration and democracy and the potential impact on society as a whole have been extensively described in the literature. Government data are expected to strengthen transparency, improve decision and policy-making processes, enhance public trust in government and democratic institutions, stimulate economic growth and innovation, and provide opportunities for developing more effective public services. Towards this end, these forms of data have been recently exploited using Artificial Intelligence (AI) methods, including machine and deep learning, to create models for prediction, classification, forecasting and clustering, among others. This special track focuses on implementing and assessing AI technologies in the public sector, exploiting eXplainable Artificial Intelligence (xAI) methods to mitigate unintended implications and build public confidence.


Enhance transparency and trust using xAI in public policy-making and deliberative democracy
User-centric XAI methods and user-friendly visualizations for public political decision-making
Explainable argument mining in public deliberations
XAI methods for converting unstructured text into structured argumentation/deliberation data
Exploring the quality, fairness, and validity of Open Government Data using xAI methods
XAI in fraud detection in the public sector (e.g., public procurements, tax evasion, smuggling)
Explainable graph neural networks for interaction modelling with public data (textual: contract descriptions/bidder info; numeric: bidding amounts/dates; categorical: types of goods/services modelling)
Enabling explainability in AI-based surveillance and predictive policing
XAI methods for surveillance data ML models (e.g 3D-CNNs, spatiotemporal RNNs with Layer-wise Relevance Propagation, Grad-CAM, Temporal Class Activation Maps methods)
Fostering transparency with XAI in contextual public sectors (e.g. prioritise cases of inspections, predict reoffending of criminals, decision on immigration cases)
XAI in public budgeting plans and resource allocation (e.g. counterfactual explanations for allocation plans)
Declarative process mining, active reinforcement learning for explainable conformance checking/diagnosis in public services provision