Table 4 Results for the discrimination between environmental-or experimental-associated sentences for different classifiers in the Weka package.
From: EnvMine: A text-mining system for the automatic extraction of contextual information
Method | Subset | Original | Classified | Correct | Recall | Precision | F-value |
|---|---|---|---|---|---|---|---|
Naive Bayes Multinomial | Exp | 828 | 802 | 755 | 91.2% | 94.1% | 92.6% |
Naive Bayes Multinomial | Env | 323 | 349 | 276 | 85.4% | 79.1% | 82.1% |
Naive Bayes | Exp | 828 | 692 | 679 | 82.0% | 98.1% | 89.3% |
Naive Bayes | Env | 323 | 459 | 322 | 96.0% | 70.0% | 81.0% |
Bayes Logistic Regression | Exp | 828 | 796 | 746 | 90.0% | 93.7% | 91.8% |
Bayes Logistic Regression | Env | 323 | 355 | 273 | 84.5% | 76.9% | 80.5% |
Bayes Net | Exp | 828 | 617 | 608 | 73.4% | 98.5% | 84.1% |
Bayes Net | Env | 323 | 534 | 314 | 97.2% | 58.8% | 73.3% |
Meta Bagging | Exp | 828 | 752 | 708 | 85.5% | 94.1% | 89.6% |
Meta Bagging | Env | 323 | 399 | 279 | 86.4% | 69.9% | 77.3% |
Rules, Decision Table | Exp | 828 | 1041 | 809 | 97.7% | 77.7% | 86.6% |
Rules, Decision Table | Env | 323 | 110 | 91 | 28.2% | 82.7% | 42.1% |
Random Forest | Exp | 828 | 776 | 735 | 88.8% | 94.7% | 91.7% |
Random Forest | Env | 323 | 375 | 282 | 87.3% | 75.2% | 80.8% |