Research Article

Performance Assessment of Multiple Classifiers Based on Ensemble Feature Selection Scheme for Sentiment Analysis

Table 6

Results of the proposed model for kitchenware review data set.

MethodClassifier
SVM MNBK NNME
PrecRecallF-ScorePrecRecallF-ScorePrecRecallF-ScorePrecRecallF-Score

Unigram85.480.282.7182.180.781.3983.981.482.6384.081.282.57
Bigram85.684.184.8481.478.880.0782.578.980.6581.283.482.28
Unigram+88.786.287.4386.384.985.5984.683.283.8983.384.583.89
Bigram
ComopIG87.985.686.7386.883.184.8684.782.683.6385.183.784.39
CompCHI86.885.985.3484.082.583.2481.183.282.1384.485.584.94
CompGI83.282.1083.1386.283.884.9885.487.786.5385.287.286.18
Average85.7583.1185.4084.4682.383.3583.8481.1882.4883.8684.2584.04