Table 2 Number of posts per domain, data source and label type for train (80%), test (20%) and overall
From: Classifying patient and professional voice in social media health posts
Domain | Data source | Patient voice | Professional voice | Other |
|---|---|---|---|---|
Train (80%) | ||||
Cardiovascular | 51 | 141 | 4307 | |
2665 | 124 | 3889 | ||
Skin | 1264 | 179 | 1835 | |
5604 | 46 | 3416 | ||
Test (20%) | ||||
Cardiovascular | 13 | 35 | 1075 | |
666 | 30 | 972 | ||
Skin | 316 | 45 | 457 | |
1400 | 11 | 854 | ||
All | ||||
Cardiovascular | 64 | 176 | 5382 | |
3331 | 154 | 4861 | ||
Skin | 1580 | 224 | 2292 | |
7004 | 57 | 4270 | ||