Article contents
From Analysis to Generation: Semiotic Prompt Engineering for AI-Generated Educational Visuals
Abstract
Engaging students and developing intercultural literacy are two key benefits of using Educational Visuals (EVs); however, the selection of an EV is often driven by convenience rather than sound methodological processes. This research moves beyond the analysis of existing representations of Moroccan culture through the creation of a new Semiotic Prompt Engineering (SPE) framework. The current study uses Gunther Kress and Theo van Leeuwen’s (2006) multimodal discourse analysis framework as a lens within the generative reality of AI models to examine the systematic curation of EVs representing educational effectiveness and cultural relevance. The proposed methodology consists of feeding an AI model a particular taxonomy and parameters that correspond to three metafunctions relative to representation, interaction, and composition, and shifting from static and independent representations to dynamic and narratively rich ones. The findings of this conceptual proof-of-concept with illustrative comparative renderings suggest that visuals generated through semiotically informed prompting techniques may portray greater depth of narrative and offer a more culturally grounded representation than visuals produced through conventional prompting techniques. In addition, the research invites a re-conceptualization of the roles of educators and curriculum designers from active collectors of EVs to knowledgeable and creative EV content creators.
Aims & scope
Call for Papers
Article Processing Charges
Publications Ethics
Google Scholar Citations
Recruitment