Verso una classificazione funzionale dei prompt basata sulle intenzioni cognitive esplicitate dagli studenti nell’interazione con l’IA

The contribution proposes a theoretical and methodological framework to interpret the prompt as a linguistic formulation that makes the function of a request observable within interactions with generative artificial intelligence systems. Moving beyond both a purely technical view of prompt engineering and classification attempts based on Bloom’s taxonomy, the study introduces a functional perspective focused on what the student does through the prompt during the interaction. Within this perspective, prompts are analyzed not according to the cognitive level of the task, but in relation to the function that the request performs within the conversation (e.g., clarifying the task, exploring ideas, requesting evaluation, or progressively refining an output). This leads to a classification of prompts in student–AI interactions based on functional categories, which can be used to analyze how generative AI is employed in learning processes in educational contexts. The contribution also outlines the design of a subsequent empirical phase aimed at testing the proposed framework