Memetic swarm clustering with deep belief network model for e‐learning recommendation system to improve learning performance
Maganti Venkatesh, S. Sathyalaksmi · Concurrency and Computation Practice and Experience · 2022
Abstract In this article, an efficient clustering and classification model using memetic swarm clustering and deep belief network called MSC‐DBN model, to derive a recommendation system for web‐based learning. In the proposed method, the three main phases are considered: (i) clustering, (ii) classification, and (iii) recommendation. In first phase, MSC uses ant bee colony, Particle swarm optimization and k‐MEANS clustering. In the second phase, DBN is used for classification in the clustered based users. Clustered based user's has three categories of learners. Finally, MSC‐DBN system is automatically recommends the learning materials to the learners based on the complexity level of the material and learner's capability. A detailed experimental analysis is conducted to ensure the effective recommendation performance. The proposed method is activated in Java. The proposed method attains higher precision 93.26%, 96.12%, and 91.11%, higher recall 89.56%, 92.36%, and 95.66% and higher F‐score 93.12%, 97.14%, and 92.11% compared with the existing approaches, like generalized sequential pattern (GSP) model with context aware (CA)‐collaborative filtering (CF) called (GSP‐CA‐CF), context aware (CA)‐collaborative filtering (CF) called (CF‐CA), and generalized sequential pattern (GSP). Finally, the simulation outcomes demonstrate that the proposed method can be able to find the optimal solutions efficiently and accurately.