International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 175 - Number 17 |
Year of Publication: 2020 |
Authors: Khadija Shazly, Marwa Eid, Hanaa Salem |
10.5120/ijca2020920683 |
Khadija Shazly, Marwa Eid, Hanaa Salem . An Efficient Hybrid Approach for Twitter Sentiment Analysis based on Bidirectional Recurrent Neural Networks. International Journal of Computer Applications. 175, 17 ( Sep 2020), 32-36. DOI=10.5120/ijca2020920683
Many optimization problems from various applications have been solved by many algorithms such as Grey Wolf Optimizer (GWO), Genetic Algorithm (GA), Whale Optimization Algorithm (WOA) and Particle Swarm Optimization (PSO). Sentiment Analysis (SA) is used to evaluate the polarity of reviews. In SA, feature selection phase is an important, as the best way to solve all optimization problems not happen yet so this paper proposes a hybrid approach that combines three modified hybrid algorithms [ (GWO), (PSO) and (GA) ], its name (HWPG) .To reduce the search space filter features selection, Information Gain (IG) has been used. (HWPG) used to select the best features for training Bidirectional Recurrent Neural Networks (BRNN) classifier. Arabic benchmark dataset which 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 (HWPG)