3rd TAGRA

This study examined software projects completed and presented in the Computer Programming program of a vocational school in terms of the problems they addressed, their target users, and their artificial intelligence functions. This descriptive document analysis evaluated the final reports of 29 software projects deemed suitable for examination. The reports were classified by problem domain, target user, the status of the artificial intelligence function within the existing product, and the described technical mechanism. Using a rubric adapted from the Conceive–Design–Implement–Operate (CDIO) framework, we also coded how well the reports described the problem, user, design, and implementation. The projects primarily focused on business, commerce, and food services (n=6); education and learning (n=5); and healthcare services (n=4). Eight reports presented the artificial intelligence function as part of the existing product, while two reports presented it as a future development. Among the eight reports describing an existing function, four detailed model or service integration, three explained threshold or matching rules, and one did not specify a technical mechanism. Using coding adapted from CDIO, 21 reports explicitly described the problem, 22 the target user, 26 the design, and 27 the implementation. The "operate" (usage) phase was not coded because this analysis did not include records of presentation evaluations. The findings indicate that students developed software projects addressing diverse user needs and employed various technical mechanisms for the functions they labeled as "artificial intelligence." For future projects, students should clearly define the problem they aim to solve and the target user, and explain using concrete examples how user needs are reflected in design decisions and the functions developed.