The HARMONIA project aims to develop methods for using large language models (LLMs), enhanced with knowledge derived from behavioral and statistical data, to support informed and inclusive public decision-making processes.

The project focuses on creating trustworthy artificial intelligence to support policy-making—from listening and decision-making to communication—guiding citizens and authorities toward new behaviors.

Although LLMs offer powerful natural language understanding capabilities, they remain insufficiently reliable for public decision-making processes. Their responses can be generic, biased, or even incorrect—especially when dealing with underrepresented communities or lesser-known situations. This limits their usefulness for local authorities, who need to understand the diverse needs of real people.

HARMONIA addresses these challenges by enhancing LLMs with structured semantic knowledge derived from real behavioral and statistical data.

Through neuro-symbolic approaches—combining ontology-based models with advanced AI techniques such as Retrieval-Augmented Generation (RAG) and Knowledge Injection—the project is designing models capable of reasoning in context and adopting multiple perspectives. The goal: more inclusive, transparent, and reliable artificial intelligence tools capable of strengthening the relationship between citizens and public institutions.