Understanding Twitter Sentiments with CNN and Gorilla Optimization

In today's world, figuring out sentiments from tweets can be tough, especially with issues like short text, abbreviations, and spelling errors. Traditional methods struggle with these challenges, so we need new approaches. Tweets often hide people's emotions, like fear or anxiety, stemming from early experiences. This study explored sentiment analysis of tweets using a Convolutional Neural Network (CNN) optimized by the Enhanced Gorilla Troops Optimization Algorithm (EGTO).Two datasets from SemEval-2016 were used to test the system. The results? The model was extremely accurate, with scores of about 98% for accuracy and 95% for precision regarding positive tweets. For negative tweets, it was almost as good, with 97% for accuracy and 96% for precision. This new model outperformed other methods, showing it could efficiently tell if a tweet was positive or negative.