Table 7. Average of proportional performance results on categories with original dataset.
| Algorithms (Accuracy) | Categories | |||||
|---|---|---|---|---|---|---|
| Peer interactive communication and technical support | Declaration | Presentation | Financial | General acknowledgement | General statement | |
| CNN (80.00%) | 8.57 | 17.14 | 11.43 | 7.14 | 21.43 | 14.29 |
| RNN (55.80%) | 5.48 | 8.27 | 19.50 | 12.57 | 6.07 | 3.92 |
| CNN+Doc2vec (93.58%) | 21.66 | 17.21 | 9.73 | 21.98 | 11.38 | 11.61 |
| kNN (79.67%) | 14.04 | 13.77 | 11.01 | 18.31 | 10.65 | 11.89 |
| Logistic regression (89.54%) | 18.72 | 15.37 | 11.84 | 20.88 | 10.60 | 12.12 |
| Naïve Bayes (88.21%) | 20.28 | 15.97 | 11.75 | 21.02 | 8.86 | 10.33 |
| N-gram language model (87.88%) | 20.24 | 15.92 | 11.75 | 21.06 | 8.81 | 10.10 |
| Average of performance per category | 15.57 | 14.81 | 12.43 | 17.57 | 11.11 | 10.61 |