Table 1.
SVM-based Sentiment Analysis Research Summary.
Fig 1.
PSO-based Global Optimization Approach for Multi-Polarity Sentiment Analysis (PSOGO-Senti).
Table 2.
Performance Comparison of PSOGO-Senti and Benchmarks in the Two-Polarity Sentiment Analysis (Ctrip Dataset).
Table 3.
Performance Comparison of PSOGO-Senti and Benchmarks in the Three-Polarity Sentiment Analysis (Ctrip Dataset).
Table 4.
Performance Comparison of PSOGO-Senti and Benchmarks in the Five-Polarity Sentiment Analysis (Ctrip Dataset).
Table 5.
Comparison between the Optimal Feature Subset Selected by PSOGO-Senti and the HowNet Dictionary.
Table 6.
Some of the Features in the Optimal Feature Subset of PSOGO-Senti but not in the HowNet Dictionary (Ctrip Dataset).
Table 7.
Performance Comparison of PSOGO-Senti and Benchmarks in the Two-Polarity Sentiment Analysis (Guahao Dataset).
Table 8.
Performance Comparison of PSOGO-Senti and Benchmarks in the Three-Polarity Sentiment Analysis (Guahao Dataset).
Table 9.
Performance Comparison of PSOGO-Senti and Benchmarks in the Five-Polarity Sentiment Analysis (Guahao Dataset).
Table 10.
Some of the Features in the Optimal Feature Subset of PSOGO-Senti but not in the HowNet Dictionary (Guahao Dataset).
Table 11.
Performance Comparison of PSOGO-Senti and the GSM-, GA-based Approaches (1).
Table 12.
Performance Comparison of PSOGO-Senti and the GSM-, GA-based Approaches (2).