Metaheuristic-based adaptive curriculum sequencing approaches: a systematic review and mapping of the literature

Abstract

The presentation of learning materials in a sequence, which considers the association of students’ individual characteristics with those of the knowledge domain of interest, is an effective learning strategy in online learning systems, especially if related to traditional approaches. However, this sequencing, called Adaptive Curriculum Sequencing (ACS), represents a problem that falls in the NP-Hard class of problems given the diversity of sequences that could be chosen from ever-larger repositories of learning materials. Thus, metaheuristics are usually employed to tackle this problem. This study aims to present a systematic review and mapping of the literature to identify, analyze, and classify the published solutions related to the ACS problem addressed by metaheuristics. We considered 61 studies in the mapping and 58 studies in the review from 2005 to 2018. Even though the problem is longstanding, it is still discussed, especially considering new modeling and used metaheuristics. In this sense, we emphasize the use of Swarm Intelligence and Genetic Algorithm. Moreover, we have identified that various parameters were considered for students and knowledge domain modeling, however, few student’s intrinsic parameters have been explored in ACS literature.