Nickel mining in Raja Ampat has triggered significant public reaction, particularly on social media, due to its environmental and social impacts. However, public opinion on this issue has not been systematically analyzed. This study aims to examine public sentiment toward this issue using the Aspect-Based Sentiment Analysis (ABSA) approach with four classification algorithms: Support Vector Machine, K-Nearest Neighbor, Naïve Bayes, and Random Forest, all optimized through the Particle Swarm Optimization (PSO) method. Data was collected from X between June 1 and June 30, 2025, and analyzed based on three main aspects, namely environmental, social, and economic, with a total of 4,025 datasets. The analysis shows that negative sentiment dominates over positive sentiment, with the environmental aspect being the main focus, especially regarding coral reef damage and marine pollution. Among the four models used, the optimized Support Vector Machine algorithm achieved the highest performance with an accuracy of 87.5%. These findings are expected to serve as an evaluation for the government regarding mining permits to formulate policies that support the achievement of SDG 14 (Life Below Water) and SDG 15 (Life on Land).