Dr Beyza Yaman, a researcher at ADAPT in Trinity College Dublin, has been awarded a Research Ireland Public Service Fellowship to explore how generative AI and knowledge graphs can improve access to decades of institutional knowledge held by the Office of Public Works (OPW). The fellowship will support a project entitled Creation of Knowledge Repository with GenAI and Chatbot Capability, examining how artificial intelligence can help the OPW organise, connect and retrieve information from its extensive digital archives. The award was announced today (10 November) as part of Research Ireland’s largest Public Service Fellowship programme to date. This year, 41 researchers including fellow ADAPT member Prof Maria Grazia Porcedda, have received funding totalling €3.9 million to work with 24 government departments and public sector organisations.
The OPW holds more than 30 years of project and operational information relating to buildings, sites, parks, monuments and other public assets. Much of this accumulated knowledge has the potential to support current operations, future planning and policy decisions, but accessing and connecting information across extensive historical records can be challenging.
Working with the OPW, Beyza will investigate how generative AI, knowledge graphs and graph databases can make this information more accessible and useful. The project will explore the development of an AI-powered chatbot that would allow staff to interrogate historical records through a conversational interface, while identifying connections between projects, people, assets and past decisions. An important element of the research will be ensuring that information can be accessed responsibly, with appropriate safeguards to protect sensitive and personal data. The project is expected to deliver a working chatbot for exploring historical OPW records, alongside knowledge graph and graph database demonstrators that reveal relationships across the organisation’s extensive information holdings.
Beyza is a postdoctoral researcher and data and knowledge engineer at ADAPT, specialising in semantic web technologies, knowledge graphs, data integration and data quality. Her research has included applying knowledge engineering to complex information systems, including work on interoperable data infrastructure and AI-enabled network operations. The fellowship provides an opportunity to apply this expertise to a public service challenge, demonstrating how AI can help organisations make better use of existing knowledge while supporting more informed and evidence-based decision-making. The programme aims to strengthen collaboration between researchers and the public service, bringing academic expertise into government to support innovation, policy development and improved public services.
