Feature Selection Using New Version of V-Shaped Transfer Function for Salp Swarm Algorithm in Sentiment Analysis

(1) Background: Feature selection is the biggest challenge in feature-rich sentiment analysis to select best (relevant) feature set, offer information about relationships between features (informative), and be noise-free from high-dimensional datasets improve classifier performance. This study aims propose a binary version of metaheuristic optimization algorithm based on Swarm Intelligence, namely Salp Algorithm (SSA), as analysis. (2) Methods: Significant subsets were selected using SSA. Transfer functions with various types form S-TF, V-TF, X-TF, U-TF, Z-TF, new type V-TF simpler mathematical formula are used approach enable search agents move space. The stages include data pre-processing, SSA-TF other conventional methods, modelling K-Nearest Neighbor (KNN), Support Vector Machine, Naïve Bayes, model evaluation. (3) Results: results showed an increase 31.55% accuracy 80.95% for KNN SSA-based New V-TF. (4) Conclusions: We have found that SSA-New V3-TF method highest less runtime compared algorithms