added dataset formation info
Browse files
README.md
CHANGED
|
@@ -1,3 +1,8 @@
|
|
| 1 |
---
|
| 2 |
license: openrail++
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: openrail++
|
| 3 |
---
|
| 4 |
+
|
| 5 |
+
Dataset formation:
|
| 6 |
+
1. Filtering Ukrainian tweets so that only tweets containing toxic language remain. Source of Ukrainian data: https://github.com/saganoren/ukr-twi-corpus
|
| 7 |
+
2. Non-toxic sentences were obtained from a previous dataset of tweets as well as sentences from news and fiction from UD Ukrainian IU: https://universaldependencies.org/treebanks/uk_iu/index.html
|
| 8 |
+
3. After that, the dataset was split into a train-test-val and all data were balanced both by the toxic/non-toxic criterion and by data source.
|