Table 4 Multi-mode experiments: F1 classification scores by older periods.
From: Detection of ChatGPT fake science with the xFakeSci learning algorithm
Publication Period (2014–2019) | Depression (%) | Cancer (%) | Alzheimer’s (%) |
|---|---|---|---|
xFakeSci | 80 | 92 | 90 |
Naive Bayes | 50 | 43 | 44 |
Linear SVM | 40 | 51 | 48 |
Classical SVM | 44 | 47 | 40 |
Logistic regression | 45 | 47 | 46 |
Publication Period (2010–2014) | Depression (%) | Cancer (%) | Alzheimer’s (%) |
|---|---|---|---|
xFakeSci | 80 | 94 | 89 |
Naive Bayes | 51 | 46 | 45 |
Linear SVM | 38 | 52 | 48 |
Classical SVM | 42 | 50 | 41 |
Logistic regression | 42 | 50 | 48 |
