Abstract
Adaptive learning in e-learning has garnered researchers' interest. In it, learning resources could be recommended automatically to achieve a personalized learning experience. There are various ways to realize it. One of the realistic ways is adaptive learning path recommendation, in which learning resources are provided according to learners' requirements. This paper summarizes existing works and proposes an innovative approach. Firstly, a learner-centred concept map is created using graph theory based on the features of the learners and concepts. Then, the approach generates a linear concept sequence from the concept map using the proposed traversal algorithm. Finally, Learning Objects (LOs), which are the smallest concrete units that make up a learning path, are organized based on the concept sequences. In order to realize this step, we model it as a multi-objective combinatorial optimization problem, and an improved immune algorithm (IIA) is proposed to solve it. In the experimental stage, a series of simulated experiments are conducted on nine datasets with different levels of complexity. The results show that the proposed algorithm increases the computational efficiency and effectiveness. Moreover, an empirical study is carried out to validate the proposed approach from a pedagogical view. Compared with a self-selection based approach and the other evolutionary algorithm based approaches, the proposed approach produces better outcomes in terms of learners' homework, final exam grades and satisfaction.
Introduction
The rapid expansion of e-learning platforms has led to an "information overload" problem, where students struggle to navigate thousands of available videos, articles, and exercises. Creating an ideal learning path requires balancing a student's learning goals, prior knowledge, and specific learning styles (e.g., visual vs. verbal). Mathematically, this is a multi-objective combinatorial optimization problem that is NP-hard. Traditional evolutionary algorithms often suffer from "prematuration," failing to find global optima. This paper proposes a decoupled framework that combines graph theory for conceptual sequencing and an Improved Immune Algorithm (IIA) for resource selection.
Methodology
The proposed recommendation framework is divided into two distinct phases: Knowledge Organization and Curriculum Design.
Phase 1: Knowledge Organization via Graph Theory
In the first phase, the system treats the concept repository as a directed graph. By analyzing a learner's current knowledge base and target learning goals, the system prunes irrelevant concepts to create a learner-centered concept map. A specialized traversal algorithm then converts this map into a linear concept sequence, ensuring that all prerequisite dependencies are respected.
Phase 2: Curriculum Design via Improved Immune Algorithm (IIA)
Once the conceptual sequence is established, the system must select the most appropriate Learning Object (LO) for each concept. This selection process optimizes for pedagogical factors, such as the Felder-Silverman Learning Style Dimensions.
To solve this optimization problem, the authors developed an Improved Immune Algorithm (IIA). While standard Immune Algorithms mimic biological antigen identification, they often stall in local optima. The IIA integrates "Prey" and "Follow" operators from the Artificial Fish Swarm Algorithm (AFSA). These operators allow "antibodies" (potential solutions) to either move toward superior neighboring solutions or explore the search space randomly when progress plateaus, significantly accelerating convergence to a global optimum.
Experimental Validation
The approach was validated through both simulation and empirical study:
- Simulated Datasets: Nine datasets ranging from 20 to 100 concepts (with 10-50 LOs per concept) were used to test computational performance.
- Empirical Study: 90 sophomores at the Ocean University of China participated in a Python programming course. They were divided into three groups: one using the proposed IIA system, one using other evolutionary algorithms, and a control group using self-selection.
Key Results
The results demonstrate significant improvements in both technical and pedagogical metrics:
- Computational Efficiency: The IIA reached near-optimal fitness in approximately 100 iterations, outperforming competing algorithms like TOPSIS and MCPIA, which required 400-500 iterations to reach similar results.
- Academic Performance: Students using the proposed system achieved a mean homework grade of 4.56/5.0, significantly higher than the 3.68/5.0 achieved by the self-selection group.
- Learner Satisfaction: Automated path recommendation resulted in a satisfaction score of 4.49/5.0, compared to 3.01/5.0 for self-selection, indicating a reduced cognitive load for students.
Conclusion
By decoupling concept sequencing from resource selection and hybridizing immune logic with fish swarm behaviors, this framework provides a robust solution for personalized e-learning. Future work will focus on integrating environmental and emotional variables to further refine the adaptive learning experience.
Summary updated