Abstract—Dealing with high-dimensional data has made it very challenging to find the best subset of chosen features because of the limitation in reducing the exponential growth of the search process. Text feature selection is a dimensionality reduction technique that tends to reduce the extra text features for better classification accuracy. Feature selection is an Nondeterministic Polynomial-time hard (NP-Hard) optimization. Additionally, many feature selection models overlook the interactions between or between features and the decision class. To produce the suggested Memetic Salp Swarm Algorithm (MSSA), the SSA is combined with Symmetrical Uncertainty (SU). The proposed technique intends to pinpoint the most valuable characteristics of Arabic text by decreasing computing complexity and enhancing classification accuracy. MSSA, drawing inspiration from the innate behaviour of salps and the concepts of memetic algorithms, effectively navigates the search space to locate the most optimum subsets of features. Symmetrical Uncertainty ranking enhances the selection process by precisely measuring the significance of the characteristics of the classification objective. The suggested strategy has been proven successful in Arabic text datasets through experimental findings, surpassing existing classification accuracy and feature subset size approaches.
 
Keywords—optimization, feature selection, memetic, classification, computing complexity


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).