<![CDATA[South Sumatra faces recurring forest and land fires, yet hotspot information remains difficult for lay users to interpret. This study designs and evaluates the user interface (UI/UX) of Sumsel Hotspot Monitor, accommodating a representation of AI-based wildfire hotspot prediction via the Design Thinking method (Empathize, Define, Ideate, Prototype, Test). Interviews with residents, disaster officers, and meteorological operators revealed a need for plain language and color-coded indicators, translated into a map prototype displaying illustrative output from LightGBM (spread probability) and ConvLSTM (movement direction) models, adopted as design references; their training and quantitative validation against real historical data are planned for future research. Usability testing (SUS) with 24 respondents yielded an average score of 78.54 (Grade B+, "Good"), indicating the prototype is acceptable for use. This research bridges AI-based hotspot prediction with user-centered UI/UX design, offering practical recommendations for an accessible mitigation application; empirical validation of the AI component remains necessary before full adoption by disaster agencies.]]>
Copyrights © 2024
