The distinction between P (process) and S (state) is unreliable: there is a low agreement on it and TUPA tends to make errors on them too. For example, in German, verbs "sitzen" and "stehen" can be marked both as P's and S's by TUPA. What can be done about it is
- Also report
weak_labeled in evaluation.
- Try merging the two labels (to
MR (main relation) in training.
The distinction between P (process) and S (state) is unreliable: there is a low agreement on it and TUPA tends to make errors on them too. For example, in German, verbs "sitzen" and "stehen" can be marked both as P's and S's by TUPA. What can be done about it is
weak_labeledin evaluation.MR(main relation) in training.