Generating Educational Mobile Applications Using UIDPs Identified by Artificial Intelligence Techniques

Abstract Commercial mobile devices vary in brand, size, and functionalities, but they all allow people to interact with educational applications. In software engineering, application development techniques, approaches, methodologies, and processes (e.g. BBD, FDD, RAD, DDD) are often time consuming, costly, or aim at expert developers, which implies that users outside the software development field (e.g. teachers) need great practice to become experienced application developers. This work proposes a process for generating educational mobile apps from freehand-generated images that represent user interfaces by using artificial intelligence techniques. The images’ design is based on User Interface Design Pattern (UIDP) representations. As a proof of concept, we introduce EduMatic, an application development tool. To test our process, we assessed Wiki-Comp, an application generated with EduMatic tool, along with three other external educational applications: Khan Academy, Wiki encyclopedia, and Kiwix. According to the evaluation results, Wiki-Comp outstands in functionality, usability, and performance aspects.