Optimize Machine Learning Programming Algorithms for Sentiment Analysis in Social Media

Abstract

Sentiment analysis (SA) is used to evaluate the polarity of reviews. In SA, the feature selection phase is essential, as the best way to solve all optimization problems not happen yet so this paper proposes a hybrid approach that combines hybrid algorithms [ (GWO and (PSO)], its name (GWOPS). To reduce the search space filter features selection. (GWOPS) used for training neural networks (NN) classifiers to select the best features. Data was collected from Twitter on different topics used in our experimental. The proposed algorithm is compared with three well-known optimization algorithms the experiments and comparisons result to evaluate the quality and effectiveness of the (GWOPS)