Sentiment analysis has a crucial role in understanding user thoughts, particularly in movie reviews, which influence decision-making in the entertainment industry. Platforms like IMDb provide valuable insights into audience preferences, helping filmmakers and stakeholders assess a film’s reception. However, sentiment classification is challenging due to the inherent complexity and ambiguity in textual data. This study presents an enhanced fuzzy logic-based sentiment analysis model that integrates AFINN, SentiWordNet (SWN), VADER, and TextBlob with Particle Swarm Optimization (PSO). Polarity and subjectivity scores are extracted using lexicon-based techniques and incorporated into a rule-based Fuzzy Inference System. Additionally, PSO is employed to optimize the classification process, categorizing reviews into five sentiment classes: strongly positive, positive, neutral, negative, and strongly negative. Experimental results show that the baseline TextBlob model achieves an accuracy of 70%, which improves to 76.5% after applying PSO, demonstrating its effectiveness. The findings indicate that the proposed TextBlob PSO model enhances sentiment classification performance, particularly in handling uncertainty and linguistic complexity in textual data.