<![CDATA[This study analyzes traditional granulated palm sugar (gula semut) production as an ethno-chemical system for contextual carbohydrate food chemistry learning through an AI-assisted applied ethnographic approach. Public YouTube videos and short online news features were treated as mediated field sites. From more than 200 sources, 53 eligible videos were selected and transcribed through a systematic data-processing workflow. Data analysis followed Miles and Huberman’s interactive model and Strauss and Corbin’s open–axial–selective coding, supported by a large language model to identify recurring practices and relate them to underlying chemical mechanisms. Four production stages were reconstructed, covering sap collection and initial processing, heating–concentration–early crystallization, crystallization and granulation, and drying–stabilization–packaging. The heating, concentration, and early crystallization stage emerged as the central physicochemical control window, where producers regulate fire intensity, foam behavior, evaporation, viscosity, color, and aroma to control concentration increase, boiling-point elevation, non-enzymatic browning, and the approach to supersaturation. Environmental conditions, fuel choice, cleanliness, and informal pH boundaries also function as additional control layers influencing product quality. The study generates practice-to-concept evidence maps and exemplar learning activities aligned with curriculum topics such as colligative properties, thermochemistry, reaction rates, and physical versus chemical change. These findings suggest that ethno-chemical reconstruction can support culturally grounded chemistry education by transforming local production practices into scientifically meaningful and pedagogically usable learning contexts]]>
