Literature Review of Software Architectures for Developing Artificial Intelligence-Based Mobile Adaptive Learning Systems

The widespread adoption of mobile devices has driven the development of diverse educational technologies. In this context, Artificial Intelligence-based Mobile Adaptive Learning Systems (AI-based MALS) have emerged as an efficient alternative to complement the teaching and learning process. However, the advancement of these systems is hindered by significant challenges such as insufficient software architectural support, lack of standardization, and inadequate requirements coverage and correctness. Addressing these challenges requires designing high-quality software architectures. Accordingly, research focused on the design, development, and evaluation of software architectures for AI-based MALS is critical from a software engineering standpoint. Although some secondary studies have explored related topics through literature reviews and systematic mapping, no comprehensive secondary review has identified the principal architectural aspects of AI-based MALS design. This study conducts a Literature Review (LR) of software architectures that integrate Artificial Intelligence into Mobile Adaptive Learning Systems. The main objective is to identify and compare the critical aspects of software architectures for proposed AI-based MALS in the specialized literature. The findings highlight several essential elements that software architectures for AI-based MALS should address, including requirements, quality attributes, stakeholders, user services, architectural views and styles, components, and implementation technologies. The primary finding of this study is the absence of a reference software architecture in the reviewed studies that comprehensively addresses the requirements needed to develop AI-based MALS to support adaptive learning in mobile distributed environments.