Sentiment analysis in social networks with social spider optimization algorithm

Swarm intelligence is a research area that investigates the behavior of insects or other animals herd with a collective manner of intelligence. Swarm intelligence algorithms are proposed for solving complex optimization problems in a wide range. In thesis study, a new algorithm for swarm intelligence called Social Spider Optimization (SSA) is analyzed. The SSA is based on the simulation of the spiders' behaviors in collaboration. The SSA performance, stability and sensitivity are compared with Ant Lion Optimization Algorithm (ALO), Moth Flame Optimization Algorithm (MFO), and Whale Optimization Algorithm (WOA). The success of the algorithm has been investigated using the existing benchmark functions in the literature. Furthermore was a sentiment analysis study carried out on the data obtained from the Twitter social media tool where data sharing is the most intensive. In addition to the sentiment analysis with machine learning methods, Sentiment analysis was performed on the obtained data using Social Spider Algorithm with a metasequential algorithm for the first time in the literature. Finally the success of the algorithms used for sentiment analysis has been examined comparatively with different metrics. Keywords: Optimization, Swarm Intelligence, Metaheuristic Algorithms, Social Spider Algorithm (SSA), Artificial Intelligence, Sentiment Analysis, Social Network Analysis