TTNWW - Separable Verb Detector + Multiword Detector


The uploaded file should be in FoLiA format, and should include annotations for part-of-speech and lemmas.

FoLiA file:

What do you want to annotate?
Separable Verbs
Multiword Entries


Separable Verb Detection

Separable verbs are verbs which consist of two parts, the head verb and the separable part. In the Dutch sentence "Ze gaan graag uit" the separble verb is uitgaan, which consists of the head gaan and the particle uit.

The separable verb detector uses part of speech tags to base its decisions on. It checks for every final preposition (VZ(fin)), adverb (BW(), free adjective (ADJ(vrij,basis,zonder) and singular common noun (N(soort,ev,basis...)) whether it can be combined with the verbs (WW) in the sentence.

If the combination occurs as a verb and is more likely than the parts being separate it is marked as a separable verb. Note that this is a fully automated tool that does not use any syntactic features apart from the parts of speech, so mistakes are possible.


Multiword Detection

Multiwords are sequences of words which can be considered as one unit. For instance Firstname Lastname.

Multiword detection uses part of speech tags and detects chains of the SPEC(deeleigen) tag.