111  Last view: 2024-10-0120  Last download: 2024-01-16

Clause Segmenter for Estonian

https://github.com/soras/osalausestaja

Clause Segmenter is a program that splits long and complex natural
language sentences into smaller segments (clauses). For example, the
sentence "Mees, keda seal kohtasime, oli tuttav ja teretas meid."
will be split into following clauses:

"[Mees, [keda seal kohtasime,] oli tuttav ja] [teretas meid.]"
(in the example, clauses are surrounded by brackets)

The algorithm mainly relies on punctuation, conjunction words, and
finite verb forms on identifying the clause boundaries.
For linguistic details/motivations behind the algorithm, see (Kaalep,
Muischnek 2012).