Architecture of an augmented reality application for coding and labeling UML class diagrams

This article presents an architecture and prototype of the UML-3Dfi application that includes using augmented reality technology for learning the coding and schematization of UML class diagrams. The proposal seeks to mitigate frequent errors between the modeling of UML classes and their implementation through the interaction between the design and the code. This prototype is validated in terms of its architecture and design compared with other tools to support the learning of class diagrams through the contribution of teachers who teach the UML modeling class in Object-Oriented Programming. In addition, the application's usability was measured through experience with second-level students of the object-oriented programming subject. The level of learning when using the tool is also validated through a knowledge test. The results were satisfactory in that the architecture had a high level of acceptance in terms of clarity and flexibility. On the other hand, the evaluation with students highlighted its potential to facilitate learning and usability. However, iterative improvements are required in the future to achieve a higher level of usability and, if possible, analyze students' emotions when using the tool for a more user-centered design.