As disinformation becomes more systematic and emotionally resonant, there is growing demand for tools that support journalists and fact-checkers in uncovering the broader narrative strategies underlying false claims. This paper presents an automated pipeline that leverages large language models (LLMs), text embeddings, and interactive visualizations to transform fact-check corpora into layered disinformation narratives. A key innovation lies in the human-computer interaction: users engage with the system through a visual interface that offers not just a snapshot of falsehoods but an evolving map of how disinformation coheres and spreads.
Ultimately, this paper contributes a new methodological approach for turning fact-check data into a strategic intelligence tool for journalists and researchers. The interplay between the user and machine-mediated through semantic processing and intuitive visual interfaces offers a scalable way to surface disinformation trends and narrative frames.