fetch text corpus from data.stortinget.no
This script focus on fetching a large corpus of text. The resulting dataset may be useful for training recurrent neural networks.
- encode to UTF8
- better cleaning of symbols/characters (currently about ~130 different)