Istella is participating in a project to experiment and develop technologies based on Generative Artificial Intelligence and Large Language Models (LLMs), applied to the content of the Institute of the Italian Encyclopedia Treccani.
The initiative, launched within the framework of the National Recovery and Resilience Plan (PNRR), aims to analyze and demonstrate how generative AI can support the enrichment, classification, and advanced usage of the encyclopedic heritage, fully respecting the editorial quality and authority that distinguish Treccani.

The project involves the development of software prototypes and experimental models capable of exploring new ways of interacting with encyclopedic content, harnessing the capabilities of artificial intelligence in various application areas:
– automatic text summarization for generating more concise and accessible versions of content;
– multilingual translation and cultural adaptation of texts;
– natural language answers to complex questions about concepts and encyclopedia entries;
– semantic extraction and organization of concepts to improve navigation among entries;
– voice interaction and multimodal access to information.

Thanks to its experience in Natural Language Processing and the construction of large-scale language models, Istella contributes to the design and experimentation of AI solutions by providing expertise in semantic analysis, model architectures, and training pipelines for the Italian language. In a constantly evolving information landscape, this collaboration represents a significant step towards integrating editorial tradition with technological innovation, paving the way for new forms of consultation, research, and learning based on artificial intelligence.