no code implementations • RANLP 2021 • Linda Wiechetek, Flammie Pirinen, Mika Hämäläinen, Chiara Argese
The precision of the rule-based model tested on a corpus with real errors (81. 0%) is slightly better than the neural model (79. 4%).
no code implementations • WS 2018 • Lene Antonsen, Chiara Argese