Accessing Malta’s PQs through a seamless AI-powered dashboard
Users can explore trends over time, analyse MPs’ activity and investigate connections between related PQs
An online platform, the PQ Dashboard (https://pq.ir.mt), is transforming how the public, journalists and researchers access and explore parliamentary questions in Malta.
The system supports democratic transparency by leveraging AI technologies, specifically open large language models (LLMs), to make PQs more accessible and informative.
The Speaker of the Parliament of Malta, Anġlu Farrugia, has emphasised the vital role PQs play in the scrutiny of democratic governments and the importance of ministers responding to them within a reasonable time frame.
All PQs are published on the official portal (https://pq.gov.mt). However, these are only available in Maltese, and the platform does not provide high-level overviews or thematic insights. Additionally, references within PQs to other questions are not directly linked, limiting navigability.
The PQ Dashboard addresses these limitations by automatically translating PQs into English, classifying them according to government functions (using the COFOG-99 taxonomy, a UN standard that categorises government activities into themes such as health, education and public order), and extracting relevant keywords.
This enables users to browse PQs by theme or explore frequently discussed topics. For example, since the start of 2025, most PQs have concerned general public service and have been directed to the Ministry for Transport, Infrastructure and Public Works. A significant number were submitted by MP Toni Bezzina.
Links between related PQs are also made explicit, allowing users to move seamlessly between them. Each PQ displayed on the dashboard includes a direct link to the official source, ensuring transparency and allowing users to verify the original content.
The system relies exclusively on open LLMs hosted on the University of Malta servers, ensuring that no data is transmitted externally. This highlights the utility of open LLMs while safeguarding data privacy, which is particularly important when applied to sensitive or confidential material.
Users can explore trends over time, analyse MPs’ activity and investigate connections between related PQs. The platform is a valuable resource for civic engagement, journalism and academic research.
Future developments include extending coverage to previous legislatures and introducing discourse analysis to categorise question and response types.
This project presents a use case for how open-source AI technologies can be leveraged to deliver value-added services utilising publicly available data, promote transparency and support democratic participation.
The PQ Dashboard was developed by Joel Azzopardi from the Department of Artificial Intelligence at the University of Malta.